The first problem is your source for the proportion of MSMs.
You used a NATSAL survey that you yourself said had too small a sample (FEB 5th).
You could have used the other NATSAL survey that was nearly double the size, but it found barely a half as many MSMs.
So that did not suit you.
You probably started off with an overestimate.
You then applied that proportion to the whole population.
That means you are counting in thousands of children who are not yet sexually active, and old men who are no longer sexually active.
All these are not at risk.
You thus overestimate the number of active MSMs thereby underestimating the risk factor.