Majority of acquitted rape defendants actually guilty?

I recently made the statement on a forum, that the majority of rape defendants who are acquitted, though innocent in law, will be guilty in fact.

I was asked to prove it, so here is my justification. First some figures:

1 – 80 to 90% of rapes are perpetrated by someone known to the victim.
2 – 18% of reported rapes make it as far as trial
3 – 40% of cases that make it to trial result in a conviciton
4 – There is a false reporting rate of around 3% (Recent DPP report on false rape allegations)

So for say 1000 rape reports, 180 will make it to trial. Of these, at most 36 will be “stranger” rape where the police would be required to identify the perpetrator. Lets assume that the police do a really bad job of this, and incorrectly identify the culprit in 25% of the time. So 9 cases will be against innocent men. (Note, this is the only figure where I am making an assumption – but it seems a reasonable one.)

Due to the difficulty of getting any rape case to trial (only 18%) it could reasonably be assumed that the vast majority of the 3% false allegations (necessarily without corroborating evidence) would be weeded out at this stage. However, for the sake of the argument, I’ll ignore this, and say that of the 180 cases, 3% will be false allegations = less than 6 further cases against innocent men. 

Adding the two figures, of the 180 actual prosecuted cases, at most 15 of the defendants will be genuinely innocent.

Yet only 40% (72) will be convicted. If we assume there are no unjust convictions, then the remaining 60% of cases (108) will include the 15 genuinely innocent men. 93 of the acquitted will, in fact, be guilty.