No, according to that site, they are all rapes. Not accusations. Actual rapes. I urge you to read the site again, there's no ambiguity in their intent.
But since you have now refused, twice, to get this point, I'll try a different approach. Let's use the reducto ad absurdum method.
- Of the 100 rapes (playing it your way, "alleged rapes"), 68 are never reported. Potentially, every one of them could be an actual rape, right? With 68% of women just too embarrassed or unwilling to run the gauntlet that comes with pursuing charges. Can you accept this outer limit, that all 68 could be real rapes that just aren't reported? Not that they are, just that they could be.
- Of the 32 reported rapes, in 25 cases the cops are never going to find the rapist. They try, but there just isn't enough identifying information or the woman recants on the story out of fear of reprisal, or for whatever reason. Reducto ad absurdum, all 25 of those cases could also be actual rapes, right? Theoretically speaking, they could all be.
- Of the final 7 cases, where the alleged rapist has been found and arrested, 4 cases are never going to get to a prosecutor. Evidence is too weak, guy denies everything, so the cops never forward it. Prosecutor is all over the police chief for "wasting his office's time and making his stats look bad," so the cases are dropped without being sent forward. Can you agree that this could potentially happen to this extreme limit? All four cases quite possibly real rapes, just not sent to the prosecutor because not clearly 'winnable'.
- So now the prosecutor has 3 cases, runs with all three, and wins 2. Two out of three ain't bad, right? In a court of law where things could go either way, winning more than you lose ain't bad, right? Heck, I'll even agree with you unequivocally on this point: The guy found innocent, he is truly innocent. Never happened. Justice prevailed.
But here's the point of the reducto argument: 99 of the 100 cases could be real rapes. It is entirely possible that they are almost all rapes.
You don't have to agree that they ARE all rapes. I certainly don't agree with that. But you do have to acknowledge, logically speaking, that 99 of the 100
could be real rapes.
Now, try your original argument using the statistics of this web site again. You'll discover how weak it is.