Iron Mountain probably only knows what data was owned by Open Source Storage, but not on the level of which data belongs to which Open Source Storage customer.

If so, it would be incredibly reckless for Iron Mountain to provide all of the data stored by Open Source Storage to one of Open Source Storage's former customers (PBS). Presumably that is why they wanted to go through the legal process so that someone else would be vested with the legal authority to determine what data PBS can legally access.

No idea how OSS and Iron Mountain were doing things so this is pure speculation, but if for example Iron Mountain were providing an object storage platform on which OSS then ran some secondary platform that stored all the data in one place with separate metadata to figure out what's what then it's entirely reasonable that they might not be able to easily separate it out. Not saying that'd be a good way to do it, but it's easily plausible

I do something similar with my clients backups, each individual client gets their own bucket on B2 for many obvious reasons but to sort out anything beyond what's obvious from the bucket name you need the backup software and appropriate encryption keys.