Gosh, a radio that is not perfect? Next thing we'll be discussing is gambling in Casablanca? Hi hi
Guess it won't come as a surprise that I agree here with most of what Jim, KX2T and Brian, K6BRN, wrote.
But, I wish Jim would tell us what he really thinks.

hi hi (good to hear from you Jim!)
And, a big thank you to Dave, G4AON, for giving us some rig-specific details. Thanks!
Yes, I'm usually the guy taking about the other half of the radio ---- the transmitter! But, guess I'll spend a few minutes commenting on the half that everyone else concentrates on.

Yep....I'm one of those that prefer my old 1970's tech "up-convert" radio, the Drake TR-7....debuted for sale in early 1978, when I bought my first one (bought a second one, an early 1979 model, about 20 - 25 years later)....and continued in production 'til early 1983, with a total of ~ 12,500 made...the last 18 months of production / the last ~ 2000 units, being the TR-7a models, which are identical except they came equipped with some original "options" (noise blanker and CW filter) as standard, and had an "input surge protection" circuit added, as well as having transmit mic signal input routed to one of the two "spare" RCA jacks on the rear panel, simplifying RTTY/FSK hook-ups.
This ole' venerable rig was designed in the mid-1970's....starting about the time the "C-line" came out (mid-1973) Drake's chief design engineer (Milt Sullivan, K8YDO) started to look at designing their next rig (the TR-7), and he hired the gentleman that had just designed the Heathkit SB-104, to help him and his team....and from 1974, or so, R.L. Drake with Milt and his team, went onto to create what would turn out to be a revolutionary radio, the TR-7....the IC-7300 of its day!

Yes, yes....I'm a TR-7 "fan boy", but I'm not a new-comer to the TR-7....bought one new in 1978.
And, I am a realist, I admit that by today's standards (50 years on), the TR-7's VCO's are noisy, its PTO (vfo) does drift upon start-up and the first 10 minutes or so....and, as for "modern features", it has no memories, only one vfo (needs the optional external vfo, to run "split"), no DSP, no "band-scope", etc...
But, it's transmitter is wicked-clean....will run 150 watts out 100% duty-cycle, etc...and, aside from the above old design issues, its receiver holds its own against many modern 21st Century rigs.
About the only thing that "modern" tech could improve on this venerable beast is lower phase noise VCO's (ya' know there have been many scientific / engineering improvements in the ~ 50 years since it was designed), and none other than Ulrich Rohde wrote about this upgrade regarding some other radios of this era, and if I was interested in serious CW contesting (where close-in RMDR and close-in IMD3 was critical), I'd buy a replacement VCO board and make some changes, and try it out....but I'm not much into CW, and certainly not into CW contesting, so no worries with the TR-7!

{oh....it's unlikely, but "possible", that changing the mixer and 1st IF amp to some more modern parts might help a tiny bit....but, unless the VCO's are changed, this would be a waste....especially since the TR-7's "up-conversion" uses a high-level DBM and a low-noise JFET 1st IF (48mhz) amp, producing a very sensitive front-end (MDS = -134dbm) without any RF amp / pre-amp! So, the only major issue is the 50-year-old VCO's. }
Here's just a brief description of the TR-7's receive front end, too bad some of our new/modern rigs aren't similarly designed (with new/modern components)...from the TR-7 manual:
Incoming signals from the antenna pass through a band-switched low-pass filter module. the transmit / receive antenna switching, and a band-switched high-pass filter module [and, these filters are low-loss air-wound coils and silver-mica caps....oh, and band-switched hi-pass filtering is something many "modern" ham rigs do not have, leaving even more ways for issues to occur.
]. These filters create an input bandpass filter. The limits of which are defined by the yellow numerals on the front panel BAND switch. A separate receiver and/or receive antenna can be connected in this path by removing the jumper between the EXT RCVR and EXT ANT jacks on the rear panel and making the appropriate connections.
The output of the high-pass filter is connected to the input of the Up-Converter module, along with the VLF antenna input and the 25 kHz calibrator output. The VLF antenna is connected through a 20 dB attenuator due to the fact that the input antenna filters are bypassed by this input. [this attenuator also reduces your VLF transmit output by 20db...so, 630m and 2200m operation is 20db lower, unless bypassing this internal attenuator....something I may do, as the harmonic output of the TR-7 on 630m and 2200m is within FCC spec as-is, how about that from a ~ 50 year old design....so who knows...]
Signals at the input of the Up-Converter module are mixed with the output of the synthesizer VCO to create a 48.05mHz intermediate frequency (IF) signal. Conversion is accomplished by a high-level, double balanced mixer to provide a very wide dynamic range. The output of this mixer is amplified by a low-noise, high dynamic range junction FET amplifier to insure adequate receiver sensitivity. This stage is followed by a four-pole monolithic 48.05mHz crystal filter. The purpose of this filter is to attenuate signals removed more than +/-4 kHz from 48.05mHz. thus protecting the remaining stages of the receiver from strong interfering signals.
In this manner, optimum receiver dynamic range is preserved while providing excellent sensitivity.
[Some have opined that retuning/narrowing, and/or adding more poles to this 8khz wide, 4-pole "roofing filter" would improve the TR-7's "close-in" receive IMD3 spec, which it would do....but, unless changing the VCO's, you'd still be "noise-limited" i.e. this would not do much to improve the RMDR.
As narrowing this "roofing filter" (1st IF filter), as well as adding further poles, would change the shape of noise pulses, which would require realignment (and possible redesign?) of the TR-7's excellent noise blanker (the NB-7, which was a ~ $75 option in 1978 ($90 in 1979), which is ~ $350 - $370 dollars today, just for the noise blanker!), and since changing the VCO's would go a LONG way to making the 50-year old design of the TR-7 compete quite well with 2020's design rigs, this (the VCO's) would be the first thing to upgrade...and then adding switchable narrower 1st IF filters, and a "new & improved" TR-7 would find itself floating up towards the top of "the list"....okay, it might not ever be at the top, but darn close, hi hi]
In 1979, a TR-7/DR-7 with NB-7, a couple narrower 2nd IF filters, etc. was ~ $1595 list / ~ $1450 - $1500 "street price"....That's a "street price" of about $6200+ in today's dollars!
Plus an addition $175 (in 1979 dollars) for the remote VFO...so, that's almost $7k in today's dollars, all-in!

And, have a look below, and ask yourself, how many "modern" amateur radios' manuals discuss the radio and its design like this?
Heck, how many RF design engineers discuss these things at all, anymore?
Maybe the guys at Apache Labs do, but I suspect few, if any, others....(maybe, just maybe, the boys at Elecraft...but doubtful they proceed with much that isn't something "whiz-bang" they can "sell" to the contester crowd?)
Milt Sullivan [K8YDO], et al, at R. L. Drake took great pride in the design and engineering of every system / part in the radio [TR-7]....you think "YaeComWoodFlexCraft" has even one guy/gal that even cares enough or has the smarts to do that?
Doubtful, but even if they do have someone that good on staff, are they given the time and authority to actually make a radio, noise blanker, etc., that is as good as ones made > 45 years ago, I highly doubt it.
Which is why we get the radios we get these days, fancy yes, great lab test results yes....but fun and easy to operate, hmmm, the jury is still out on that!

Here's a quote from the TR-7's manual, discussing just the NB-7 Noise Blanker (and, this is just ONE circuit):
Circuit Description:
This noise blanker system is comprised of the three major networks described below. Refer to the proper schematic for your particular version to follow this description.
Transmit Path
In transmit, diode CR815 is turned on with +10T via RFC812 and RFC813 from pin 37. The 5.645 MHz double sideband transmit signal is fed to the output coax connector through C833, CR815 and C838. When CR815 is on, CR814 will be reverse biased, thus holding the receive path off.
Receive Path
In receive, diode CR814 is turned on with +10R via RFC810 and RFC811 from pin 24. In version 1, the receive signal is applied to pin 22 and coupled directly to the blanking gate, comprised of T810, CR812, CR813, and T811, then through C830, CR.814 and C838 to the output coax connector. In version 2, the receive signal again enters from pin 22, however, then passed through a matching amplifier consisting ofQ816 and associated circuitry. The output of Q816 is then coupled to the blanking gate of T811, CR812, CR813 and T811, passes through C830, CR814 and C838 to the output coax.
Noise Processor
The Noise Amplifiers consist of Q810, Q811, and U810 cascaded and tuned to 5.645 MHz by L810, L811 and L812 respectively. The output of the noise amplifier string is split by C828 to the pulse detector and C827 to the noise amplifier AGC circuit. Q812 and associated circuitry comprise the noise amplifier AGC detector and amplifier. The AGC voltage is applied to gate 1 of Q810 and Q811 via R826 and R829 respectively.
The pulse detector, CR811, responds only to the positive half of the amplified bipolar input pulse. The network of R839, C831 and C835 wave shape the pulse at the base of the pulse amplifier QB 13. Again, the output pulse of Q813 is shaped by R847 andC840 and is applied to the gate driver, Q814. Resistor network R842 and R843 provide fixed reverse bias for the blanking gate. Q815 is a DC switch for +10R and +10NB.
Theory of Operation
The 5.645 MHz receive signal, with noise pulses, is applied to pin 22. In version 1 this signal is coupled directly to the blanking gate. In version 2, amplifier Q8I6 amplifies the signal and noise pulses to drive the blanking gate. Tuned amplifiers Q810, Q811 and U810 amplify this low level signal up to a high level to drive the pulse detector CR811.
This detector responds only to the positive going portion of each noise pulse from the output of U810. Following the detector is an RC network which shapes the pulses for driving the level shifter Q813. Again, on the output of Q813 is still another RC network for wave shaping. The gate driver transistor Q814 responds to the negative going pulse from Q813 which allows the blanking gate to tum off, thus muting the receive path and blanking the noise pulse.
Since the noise amplifiers run such high gain, Q812 and associated circuitry comprise an AGC loop to maintain a near constant output level to the detector. This allows detection and processing of very weak as well as very strong noise pulses without degrading the blanking action.
Now, some are saying...."huh? What is this guy spouting off about? This is a thread regarding a perceived issue with a FTdx-10 image issues when operating way out-of-band, why is he off on Drake TR-7 fan-boy advertisement?"
Well...
Well, the answer is....to show you all that there is no "perfect radio", and to try to compare one designed for one purpose / in one era, to others designed for different purpose/different era, is sort-of a waste.

hi hi
And, Brian went to the Kenwood 440 vs. the R-5000, to make a point, and Jim compared the Yaesu's to his old '7300 and '7610....so, I'm just a bit more long-winded.
But, if some also gain a new respect for an almost 50-year-old TR-7, well that's a nice secondary plus.

73 to all,
John, KA4WJA