Yes. As a referee, the normal process is often something like: read paper, think about paper, have questions, ask questions in the report, suggest R&R, if questions are answered in responses, then suggest accept. Under the "no revisions" policy it's hard to adapt.

Tried it once, didn't go through. I have a couple of pubs there through the normal process. Also have had lots of conversations with a previous editor about it. I think it's a great idea, but the problem is that it changes the equation for referees in a way that they don't entirely internalize.