Twitter let the fake accounts grow for years while they were publicly traded and unprofitable, because then they could at least claim they were investing in growth (that the product was too hard to use but they were trying to fix that problem and eventually these "eyeballs" will be monetized).
Now that they've been profitable 3 quarters in a row they can keep pruning spam. They don't need to show the fake growth now that they have actual real profits.
They've been banning
over a million accounts per dayThe thing the press missed was their active users were up 11%, so it was mostly just the garbage fake stuff getting torched by Twitter being used as justification to knock down the stock on an echo of the Facebook fears.
There's the shadow banning concept too, but that is a parallel topic to the total users or total active users or such.
If Twitter sees active users decline and/or ad revenue growth sharply fall, they'll probably allow a lot more spam accounts onto the platform again (if they can't profit sufficiently from the real metrics, may as well allow alternate metrics to carry the stock).