One commonly mentioned danger of a large corporate player like Twitter coming into the decentralized social world is the Embrace-Extend-Extinguish tactic, wherein one entity adopts a common protocol, and once it acquires a large marketshare then makes itself incompatible through custom changes, locking its users in and leaving the common protocol for dead.

Something like this happened to another federated/decentralized technology, XMPP. XMPP was used by Facebook Messenger, WhatsApp, and Google Talk to power instant messaging, then one by one those platforms disabled federation features and now barely anybody still uses XMPP. People fear something similar is going to happen to e-mail through GMail.

However, I think that in our case such a scenario is highly unlikely. First of all, the order of events is reversed--ActivityPub and Mastodon have been built after Twitter, as improvements over Twitter, and the current, not insignificant userbase has specifically left Twitter over bad features, management, or misaligned incentives, so there is essentially no risk of losing that userbase back to Twitter even if it joined the space.

A lot of what makes Mastodon's model attractive depends on smaller nodes being better, so it would not have an edge in attracting new users either. So, I think we only have something to gain.

I think Twitter adopting ActivityPub would be a good thing and a victory for the web. Twitter's current lock-in on its users is its strength, and weaking that lock-in by embracing decentralization would remove essentially all network effects-based obstacles to people moving away from it and spreading out in the fediverse.