It certainly may be really only the way the burner is tuned. Just like with any fuel, the air/fuel mixture needs to be correct to get the maximum output available from whatever oxygen level you have at any given elevation.
It sounds as if the new frig was tuned for lower elevation than the older ones mentioned. That tuning would make the burner at correct mixture at maybe 1500-1800 feet, lean at sea lever, and rich at the 3200 feet. At some point the burner gets too rich and won't generate enough heat, not light well, fouls the burner with soot, so if it is tuned to a lower elevation you get richer, sooner, as you go up and hit the bad point earlier.
No matter how you tune it, you max output is still going to be reduced by the reduced amount of oxygen available, following the same % reduction as a normally aspirated engine would, I think.
They may have tried to optimize the burner tune to the elevation they think was used the most improve cooling. Tuning for lower elevation would put more gas input into the burner, so more cooling capacity, but it will also go too rich sooner going up in elevation.
On edit, here what a quick search came up with for derating burners for elevation.
http://www.upgnet.com/PdfFileRedirec...1.13V-0496.pdf
Seems to be orifice change only, but I assume the mixture would be tweaked with the burner air inlet to reduce CO.
I saw else where to use the same 4% per thousand feet increase that is common with gasoline engines, but several places said that once the mixture is corrected for elevation, the derating of heat output would be more like 2-3%, which is also similar to what happens when you correct the mixture in a gasoline engine, which the computer does automatically these days.
Linear Mode
