This particular factlet dives into the minutiae of social security projection details a bit, but it is also a hugely overlooked detail that could entirely change the perceptions of the health of social security.
Unless things have quietly changed in the last two years (and it's quite possible; I invite correction from knowledgeable commenters), the fact is that social security projections are based off of prior economic cycles. And the social security projections define an economic cycle as "peak to peak" - meaning, from the end of a recessionary period, to the beginning of a next recessionary period.
Over the past 34 years, the last four complete economic cycles are from 1966-1973, 1973-1978, 1978-1989, and 1989-2000. Those are the four economic cycles that social security projections are based off of.
Notice what is missing - any period of time since 2000. That is because we have not had a recognized recession since then. Since we haven't had a recession, we haven't yet reached the second "peak". What's happened since 2000? We've had continued economic expansion that surpasses the average expansion of the previous 34 years. There's this little thing called the internet, see, that increased productivity. That was also true of the late 90's, but that only had a partial impact on the 1989-2000 cycle.
Since social security projections are conservative, this last period of economic data has been left out of the social security projections entirely. But now that we are quite possibly hitting a recession, it will be incorporated into the projections. Which means the projections for the health of social security will improve, perhaps by quite a lot.