I agree with you/Bostrom that those two things seem very unlikely to be universally [multiversally?] absolutely true. I think the potential issue with the argument is not those but the first assumption: there may be too many unknown unknowns to say for sure a willful(-ish) actor (or an infinity of them) would necessarily, even be at all likely, to actually lead to your being more probable for our given reality hyperobject.
It could be true, it is admittedly logical, just as it is Shakespeare's work was of a man and not infinite monkies, but the exploration space for the monkeys still potentially feels way beyond us right now and maybe forever. That's kind of just me making a vague appeal to hypothetical metaphysical meta-woo but I guess it's somewhat unavoidable for these particular categories of existential consideration
I see it all more as a "it's foolish to discount p(sim), and the 'rationalists are accidentally rediscovering theism lmao' snarker dunks indeed are painfully moronic"; like an anti-anti rather than a pro case. So I too am surprised Scott's is > 50%. (Guess I have to update mine now because what are the odds I would have ended up in a universe where Scott was wrong about an important issue I got right...)