Issues with large (as in message count) mailboxes is a common thing nowadays. Don't know about client side as I've been working on the server side for past few years (see https://wildduck.email/). Clients with large mailboxes cause a lot of issues both server implementation wise and operation wise. For example some widely used IMAP server software dies if an IMAP client tries to SELECT a maildir folder with >320k messages (every message in the folder yields in some file system calls while SELECTing, it takes some time and then the process that handles the client connection just crashes due to some overflow or whatever). Also there's older versions of Outlook (eg. 2010) that tend to sync state by FETCHing FLAGS of all messages in all folders every X minute and causing huge amounts of traffic and so on.
I'm not terribly surprised. It's easy to forget that mailboxes are actually databases, and at the scale of GB databases, you're going to start stealing ideas from the database community on how to manage databases rather than relying on the filesystem to act as a database for you.
Thunderbird itself does a nastily poor job of scaling to massive folders, since it assumes that it can do UI-thread-synchronous operations like "get a list of every message in the database." (And fixing it is hard... as I have way too much experience with).
