Well, I did it- I purchased the Champion 1600/2000 Inverter Generator for $499 at Sam's Club, and have some initial test results.
I placed the generator fairly close to my shack window (kept the window closed to prevent exhaust from coming in) so that I could listen to it while experimenting to assess the load from an audible standpoint. I connected the output of the generator to a surge protected plug strip using a heavy duty extension cord with 12 gauge wire. I then plugged the following devices, which I will need for my field operation, into the plug strip:
Samlex 1223 power supply (which powers my K2)
Small external fan for amp (500 ma Draw)
Small external fan for K2 (500 ma Draw)
Lenovo Notebook power supply (70 W)
All of these items were turned off initially. I started the generator (in normal mode- not economy mode) and let it warm up for about 1 minute, and then turned on each of the items in turn. There were no issues- the generator responded with a minor lug from the in-rush demands by the power supply and amp when they were turned on, but this is typical even with home power.
I connected a digital volt meter to the power strip, and measured the power at 124.4 VAC, with no fluctuation.
Next, I tuned up the K2 (with the amp in bypass) on 15M, and ran it up to 90 W key-down. The generator responded as expected, with a minor audible increase in loading, but no problems noted. I then reduced power on the K2, and tuned up the SB-200, starting with a low drive value to give about 150 W out.
NOTE: I made sure that all antennas/dummy load used for this testing were very well matched using a heavy duty AT-1200 manual tuner and a MFJ-207 antenna bridge prior to tuning the amp or K2. This will also be done in the field- I don't want to introduce any unnecessary inefficiencies.
I then tuned the amp up to 400 W (measured on my old HM-102 watt meter), and noted that the generator responded as expected, with audible loading. The voltage observed on my DMM dropped to 122.8 VAC- not bad considering the load. I then proceeded to make a CW QSO running at 25 WPM. I got a very good report with no noticed tone issues or other signal problems. The generator could definitely be heard responding to the dynamic load being placed on it, but that's to be expected. It never kicked off, or behaved in way that caused concern.
To increase the load, I connected a small 500 W heater to the plug strip, and ran it at the same time. The generator continued to run along with no issue, even well running CW at 400 W. I did this to make sure that I had some "head room" with available current from the generator.
I then switched to SSB, re-tuned everything, and had a couple of SSB QSOs. I received good audio reports with no complaints of signal problems. The audible loading of the generator was much lighter when running phone vs. CW, as expected. I continued testing on 20 and 40 meters, with similar results.
I should point out that I had no noise issues from the generator (hash, hum etc.) on any of the bands tested, which was a great relief.
Thus far, I'm very pleased. This generator is well built, and is very quiet. I don't have a Honda, so I can't do a side-by-side comparison, but I can say that the noise level is acceptable and I'm very satisfied.
So the real test will begin on May 7th, when I take the generator out and run for 18 hours during 7QP. I will give an updated report following that event. In the meantime, If I get the chance, I will try to rig up a cable to connect my O-scope to the output so that I can assess the AC waveform when a heavy load is applied, and see if it distorts or varies from 60 Hz. At this point, I think it's safe to assume that it's OK, based on my on-air results.