Do you actually know how long it lasts? What if it lasts only 100 years? With DNA we know the lasting period very well because of carcasses and corpses that died thousands of years ago and whose DNA we were able to sequence.

DNA's half-life is actually around 500 years. A major issue with ancient DNA samples is degradation. We have techniques for working around this, but part of that relies on knowing the underlying genome that we are sequencing (since we care about variation from that reference genome).

you’re making the assumption we’ll use the same type of DNA and that this will work in exactly the same way. pretty powerful assumption

Well, it’s impossible to have it both ways. If you want the already validated approach then you’ll have to stick with DNA and not some variant structures.

It’s a decent point, however, that if this tech takes off, there may be innovations in storage molecules to improve read/write speed, density, etc.