A bit late back to this debate.
Part of the reason for me starting this thread was an attempt at understanding some common myths which go around the GA scene.
For example, some piston pilots say they find the best MPG is around 7000ft. Yet, I cannot reproduce this well. I do indeed find that the lowest altitude at which I can fly at what should be an efficient engine operating point (wide open throttle, 2200rpm, 25LOP) is about 10,000ft, I lose a suprisingly small amount of range if I have to do a slow climb to say 18,000ft (to stay above weather) followed eventually by a descent.
I would have thought that the engine efficiency (NON turbo) would fall at the lower power outputs which are inevitable at high altitudes, resulting in a loss of range as a result of the time spent high up, but this doesn't seem to have much of an impact in practice. I have a GPS-linked fuel flowmeter which gives me a forecast reserve at destination; accurate to better than 1%.
If I take the Lyco HP v. fuel flow graph and extrapolate the (say) 2200rpm line all the way back to the 0HP intercept, there is a significant fuel flow there, a few USG/hr, which I guess represents the friction and pumping losses at 2200rpm. These, I would have thought, were constant at that rpm, but maybe the pumping losses decrease with altitude because the air is thinner and flows in more easily? Also, at altitude there is less exhaust back pressure so the engine will be more efficient.
There are so many effects all working at the same time.
One thing appears certain: roughly speaking, at a constant thrust, you get a constant TAS. So, as one climbs, the IAS obviously has to fall (which it does) even if the LOP fuel flow is kept constant.