On both these points I've been posting a bit in another thread on this based on following the published numbers each month (for a while now):
Point 1 is that I don't think the idea that everyone who gets stuck beyond, say, 16 months or so (or perhaps in the 'more time needed' basket) 'will' end up waiting 24-30 months for security to get to it. As far as can tell, it just can feel like it because a) at that point there are not many left in the monthly cohort, b) the rate of files approved/cleared is not very high on a monthly basis, although distinctly non-zero), and c) we don't see a lot of posts here / other fora - partly because of b (i.e. there's not a lot to report), partly because any one forum is a smallish subset of the cohort, and partly because people tune out to save their mental health.
In other words: it's easy to get the impression that no files are being cleared/approved during the 'more time needed' period. But it seems to be fairly steady, if slow, and it's just that once the numbers in a monthly cohort fall below, say, 15% of the original, the number cleared each month is small.
And yeah - weird that the three months Jan-Mar 2025 flipped to more time needed last month and back to 1 month. I don't really have a good explanation for that, except for two very simple explanations:
1) The monthly estimates are absurd on their face (there are currently fully seventeen monthly cohorts with an estimate of one or two months processing remaining).
2) Somewhat natural implication of a distribution curve of processing that's (roughly) a normal distribution (what we'd often call a bell curve) but with a very strong skew to the right, or a 'long tail'. Meaning, a large chunk of files from each month get processed fairly quickly (5-10 months) and after that, the rate falls to a low rate. [Which to be clear, most of us already knew] Note the 1-2 months remaining group currently covers ... all of the cohorts currently more than 10 months after submission.
Now their 'how many months left' having what appears to be farcical implications (including three cohorts going back and forth from 1 month to more time needed and back) could be either some very sophisticated estimate of probability based on decay rates and thresholds with some really complex statistics that's sensitive to some parameters or ... an incredibly simplistic and mechanistic approach that overall makes not a lot of sense but gives a number people sort-of understand and gives the desired result - fewer calls to the hotline asking 'where's my file' and an answer the call centre staff can use ('1 or 2 months, maybe'.)
Oh, and it also turns out to be sensitive in unexpected ways to a few parameters.
My guess is the latter - it's a very simple formula but don't tthink about it too much, the desired result is a simple number that reduces the number of calls/simplifies the response to call centre inquiries. (And the meta-result is no-one should think about this very much, including me, but too late for me, so 'sauve qui peut.')
