Clearly W0IPL and I have some different experiences working NVIS (not to mention different ideas
how a transmission line works), though the general approaches are similar.
I recommend target heights of 20 to 25 feet - that gives you a bit more efficiency (~2dB) as well as
somewhat less variation in antenna tuning due to the ground characteristics. The low angle response
of a dipole antenna doesn't change that much over this range of heights - for example on 80m the
response at 30 degrees is about 5 +/- 1dB down from that at 90 degrees for antennas less than about
40 feet high. (Or 30 feet over very good ground.) There just isn't enough difference in low angle response
of the antenna itself compared to the high angle response to make any practical difference in the ability
to copy a local station through the noise. But the gain due to height isn't a lot, and an antenna at 10'
will still work if that is what you can manage.
A couple thoughts on the issue of higher noise with an increase in height: first, if interference from
distant stations is a limiting factor, you're probably not on the optimum band for the path and need
to move to the next lower band. But since the antenna patterns themselves don't show any
relative difference, my guess is that it is due to vertically polarized pickup on the feedline due to
common mode currents. I haven't experienced such problems (though I've certainly seen high
common mode currents with simple dipoles) but it is a good reason to use a balun at the feedpoint
of any such antenna. (Our ARES trailer has a problem with interference from the battery chargers
that causes similar problems, and a balun reduces the pickup into the receiver.)
Note that this is INDEPENDENT of whether or not you use a tuner: it is about preventing the coax
shield from being part of the antenna and picking up local noise and/or vertically polarized or low
angle signals (which may also be reduced by having less of the feedline running vertically, e.g. having
the antenna closer to the ground.) I think this is one of the biggest problems with tuning up
a very narrow bandwidth antenna such as the Hamstick Dipole: the tuning becomes highly
dependent on how the feedline is positioned.
A good estimate of antenna performance is the average height of the ends and the center: raising
either one always increases signal strength when the other remains fixed (up to some reasonable
maximum: I've only analyzed the data up to about 50' on 80m.) There isn't a lot of difference
between a vee or an inverted vee, but the losses are slightly lower with the feedpoint elevated than
with the ends elevated.
If you are operating from a base camp or other area with lots of people, this is another reason to
keep the ends of your antenna up in the air where they are out of the way. Having 12' or 16' sectional
masts available to hold the ends up makes it much easier to coexist with other users.
What I would worry about with dual 80m hamsticks is the radiation efficiency. With 100w input, one is not likely to radiate more than one watt. A 24' dipole with high-Q loading coils would probably radiate +10dB (10 times) as much RF as two 80m hamsticks.
Here is where we get to compromises, and an analysis of what you expect to need to be able to do.
With proper choice of band NVIS can work well with very low powers - certainly those of us who
operate QRP have shown this many times. As I mentioned above, I've heard signals from an FT-817
in SWR shutdown feeding a Hamstick dipole on 80m, and another station on 160m who simply used
his 80m dipole at whatever power level the rig cut itself back to. Neither was particularly strong,
but both could be copied.
If you are only operating for a short time and have 100W available, the efficiency isn't as much of
a problem. You can stick up a Hamstick dipole, use the tuner to get a match, make your contacts,
and move on. Even a 20dB loss in antenna + feedline might not prevent you from making contacts.
But if you are going to provide communications for an extended period and have to make the best
use of limited battery power, or are working over marginal paths, then improving the antenna to a
full size dipole can make a big difference, as it would allow you to use lower power (and fewer
repeats) to get your message through.
This is where you have to have a clear understanding of the circumstances you plan to be operating
in and plan accordingly. We still carry the Hamstick dipole kit in the ARES trailer (though I'm proposing
that we get some extensions to make it more efficient) because we might encounter a situation where
that is the only option. But, even in my years with Search and Rescue, I haven't seen one yet.