Yes, 51% is a statistical majority, there is no argument there. A tendency is an inclination towards something. But I don't call a 51% chance "very likely"
Your statement here:
I've not seen one example of a cartridge/caliber change that "tended to" make a firearm less reliable.
is subject to examination. There have been, for instance, many .22 conversions for let's say a 1911 originally made in .45ACP
Even if the conversion was every bit as reliably mechanically as the original chambering, the fact that .22s themselves are not as reliable as centerfire cartridges in general, means that the change tends to make the firearm less reliable in .22
The M1 rifle at one time had a conversion to .308, and some of those conversions were not reliable, in the way that the inserts were problematic. I would have to say then, that the original chambering trended towards more reliability