your argument was right until the last bit. assuming what you said about women having 5% masculinity due to common parts to be true, then a trap, who looks exactly like a woman, but has a penis, would have 5.257% masculinity. this, he would still be more gay than liking women.
but look at it this way:
let m be the percentage of masculinity of a person
let Gayness be a function of masculinity, g(m) varying from 0( straight) to 1(gay)
let ∆m be a tiny increase in masculinity, as seen in a trap.
therefore, there will be a corresponding increase in the gayness, given by ∆g
∆g = ∆m (dg/dm)
, where dg/dm is the rate of change of gayness wrt masculinity.
now, we know that g(5)=0 and g(100)=1 (by postulates of homosexuality)
therefore, dg/dm can be approximately as (1-0)/(100-5)
=1/95 = 0.0105
thus, on a scale of 0 to 100, traps would be approximately 0.3 gay, which is straight for all practical purposes.