FWIW, I've been through exactly that, and I disagree with you on most of your points.
You said "most" of my points. Which are the ones you agree with?
I also have to point out that your assertion here:
"Anecdotal data welcome, but mark it as such. It's useful to illustrate a point, but not enough to prove one."
contains a fairly major flaw. It carries the implication that the data does prove these points. It doesn't. There is nowhere near enough data, nor complete enough data (and I doubt there ever will be, for a lot of reasons) to prove one side or the other of the "homebirth is safer" vs. "hospital birth is safer" debate. The issues are far, far too complex. For example, my case , where my baby died during an attempted homebirth, but after transfer, and there were multiple factors going back to my first "birth" 14 years prior, and even to my own "birth" 39 years prior, that affected the outcome. There is no data set in the world that would show everything that went into that, and since every birth is different, the data isn't going to prove a point about which is safer for any individual woman/baby dyad.
Actually Evidence based medicine (or in this case midwifery) is a matter of statistics and probability. If one thing is mostly dangerous or safe to a few hundred moms, it is probable that it will be mostly safe or dangerous for others too.
"Proving" in this case means "having the greatest likelihood and is therefore recommendable". An anecdote is looking at just one case. Statistics look at a few hundred or a few thousand cases, so a statistic has a better chance of being "right"