Whilst I agree that universal consuming nanobots are a bit far fetched, I’m not sure I’m sold on the replication problem.
Life has replication errors on purpose because we’re dependent on it for mid to long term survival.
It’s easy to write program code with arbitrarily high error protection. You could make a program that will produce 1 unhandled error for every 100000 consumed universes, and it wouldn’t be particularly hard, you just need enough spare space.
Mutation and cancer are potential problems for technology, but they’re decidedly solvable problems.
Life only makes it hard because life is chaotic and complex, there’s not an error correcting code ratio we can bump from 5 to 20 and call it a day.
All they have to do is convince some of the scientists to peer review each other’s work for free and theres no longer any significant difference between their scam and the OG journal scam.