Bear in mind that wire capacities are often rated at a
particular temp, or in free air, and in the transformer
there is a real thermal path problem for the captive
winding layers. As current goes up, so does self heating
and the current capacity derates. How you'd figure that
accurately, is a challenge.
It's pretty easy to get a 2:1 disagreement just with tens
of degrees of assumed temperature. And if the temp rise
allowed is more about insulation degrading, than copper
fusing, that's a whole 'nother dimension of concern (and
likely a well lower max use temp),