There is a huge amount of what I think in unawarded debate about applying it to the past (IMO, it's very clear where it should be used), but applying it to the future is completely against any kind of logic.
It can certainly lead you to some pretty wacky-sounding conclusions (looking at you Frank Tipler) but I can’t see why it’s obviously wrong for future observers to deploy it in just the same was as we do.
You can't apply conditional probabilities to unbias some result if you don't know if the condition holds or not.
You can invent some fiction where some future people apply it. You just can't pretend that there is any realism on it.