I'm with Babcock Hall on this one. I've had many discussion with Souro who advocates gas phase acidities. I understand why this could be good data, but I think it has a different set of issues, so we come back to why does this difference exist.
If we go back to some simple principles that are consistent with the question. If one compares ammonia, ethylamine, and diethylamine, one would have concluded that carbon is more electron donating than hydrogen. However, triethylamine does not follow that trend. Dimethyloxonium cation is a stronger acid yet. You may find various arguments, but I'm more inclined to think it is related to solvation properties (though I cannot quantitate the actual effect). I'm not bothered if they affect oxygen differently than nitrogen.