Bears are fairly territorial, and even when they abandon territoriality because of a particularly rich feeding ground, you're not meeting dozens of new bears every day unless you're hiking along the length of a river, which is not how the bear naturalists typically operated (and furthermore, they're rarely in the "meet a bear in the woods" range). Still, that uncertainty is why I put the range at 100-10000.

Your personal account of having to escape robbery is saddening--I'm sorry you've been in danger and that people you know have been assaulted. I did look through quite a few trail runner accounts (~20) without finding any accounts of substantial danger, but the people who haven't had any issues may be biased towards writing glowing accounts of the benefits of it and/or may have decided to elide some bad experiences. These were some of the less good estimates because although murder is pretty widely reported, any lesser assault (sexual or not) suffers from under-reporting, so the guess was pretty wild on that side.

I had neglected robbery because I judged that rape is worse than robbery, but also probably more likely "in the woods" at random because mugging is quite rare as a spontaneous crime. However, if you talk about which men are actually just-barely-in-the-woods, the number of muggers might be increased enough so you can't discount it. It would depend how you encounter your men and bears: do you go to them, or are they brought to you?

I haven't seen statistics on what robbers perceive, but I have seen statistics on what they manage to do, and the stats I've seen indicate that they're slightly more likely to use violence against women, but considerably more likely to attempt to rob men. So I don't think we know how this plays into any calculation. Are women more cautious, and fail to give robbers the chance (at which point the bear vs man calculation removes their chance to be cautious)? Or are there enough robbers who have internalized the fairer sex stuff (or simply just like women so wouldn't want to hurt or rob women who are random strangers) so that they'll go for (more dangerous) men because they don't have to feel as bad about themselves that way (in which case, you might draw one of these men with bear vs man)? Hard to tell.

And you're quite right that the people you find in random places in the woods aren't the same as those you find on trails in the woods (for instance), so, absolutely, the statistics from things like national parks might not be appropriate. I just don't have any good way to tell. I'm picking the best sample I can for which there's any data, which is at least a lot better than murder rates dominated by urban crime.

Like I said, it's not a very good account. My estimates vary by factors of 100. It's just that everything else I've seen has been so much worse that it confuses more than clarifies.