I think you're a bit too data-driven; where data is not present, logic suffices. The absence of data is not justification for suggesting we cannot make claims that are meaningful and weighty. That philosophical note aside, here are my replies, such as they are.
What is your explanation or evidence to support a claim like this?
Logic. If users do not have to write RFDs, it becomes easier for them to vote bomb (less effort involved) and harder for moderation to detect (sense we cannot draw inferences from their RFDs). Instead, moderation would need to rely more heavily on patterns of voting, which remains our chief tool in the status quo, to make those kinds of assessments.
I don't think you can say most people won't take the opt-out debates until you have data to support the claim.
I do have data, though. DDO has an opt-out system. No one used it. Like, literally no one. Besides, most people are going to want to be able to appeal to have obvious vote bombs and bad votes removed, because most people aren't really going to want those kinds of votes to stand on their debates. Have you seen how pissed people get when they report votes and those votes don't get removed? Imagine that, but multiplied. So, no, most people are not going to opt-out.
Which I think will happen more often than not but I have anecdotes to support that is occurring right now when they didn't know the restrictions put in place by the instigator.
I don't think it will happen "more often than not," but I do think it will happen "more." Of course, someone might say that it's a case of caveat emptor. If you accept a debate without reading the rules, someone (not me) might argue that you can't complain when the rules come back to bite you. It's a legitimate argument. Certainly, it's mitigating, but I don't think it totally defeats the argument you're making.
I don't see why a less rules option is even being democratically voted on. From my eyes it just gives rise to mob voting and creating mobs.
It's one thing to have trials, which create systems of mob rule that can hurt users who never consented to those systems. But for opt-out debates, since both debaters presumably consented to the consequences of opting out, there is an argument to be made that there is no harm in the policy. Either you consented, and so you can't complain. Or you failed to actually read the rules, and so you can't complain. Either way, you can't complain. That would be the argument for the policy change. Mob rule implies tyranny of the majority, but that cannot exist where the minority consented, and, in the case of the opt out, the assumed minority (the debaters) would have consented.
I am trying to present arguments for both sides. As I said earlier: you're weighing the mitigated danger of users inadvertently accepting opt-out debates against the benefits of increased flexibility for that minority of users who wants it. That's really the choice that's out there.
