Do these surveys also interview the alleged rapist and ask them their opinion about whether a rape has been committed?

Generally you don't track them down, no. That would be a lot more expensive study that a random survey.

We have this wonderful Anglo-Saxon tradition of juris prudence where you're innocent until PROVEN guilty.

This is not part of the justice system.

In other words you asked the other guy to disprove your evidence free assertion of bias

Either there is or there isn't. There's anecdotal evidence that race is a barrier to reporting rapes, as I have linked to.

It needs evidence on both sides. But he hasn't even shown his evidence for "In almost every case, the race of the victim and the perpetrator are the same." Yet, so lets leave the discussion of reporting bias until we understand the data he's referring to, as that will affect its relevance.