Note: I haven't run this response by my collaborator Jonah, so he may not agree with some aspects of my answer.

I guess you're concentrating on "Reason #1: Downward update on social value"

As we mentioned, one reason for the downward update on social value for us was that we though we're already close to having plucked the low-hanging fruit with respect to receptive advisees. In other words, it was an update in the direction of "the target market is smaller than we thought." We haven't updated much in terms of our view of per capita impact.

I think that in the case of both CFAR and 80K, they are okay with a relatively small target market because they have a mechanism for generating huge returns per audience member: CFAR can charge people for workshops(something we can't do -- our advisees are young and cash-constrained students who would generally have been unwilling to contact us if they had to pay for it). 80K hasn't charged advisees (as far as I'm aware) but they get huge altruistic returns per capita by affecting people's career decisions in the immediate aftermath of the advising process.

We were aiming to affect people's general way of thinking and learning while they were students, perhaps several years before they join the workforce or make other direct contributions. Since there's a greater time lag between people being exposed to our ideas or advising and actually getting out into the workforce, there's a greater level of fade-out -- of the people whom we reach out to, only a small fraction will apply it to their lives a few years down the line. Given this damping factor, reaching out to a larger audience was more critical for Cognito compared to 80K and CFAR, and our pessimism about whether we can do so was one of the reasons we decided to switch to maintenance mode.

Another point is that both 80K and CFAR started out as parts of more established organizations/centers, therefore they have had more time to develop and grow before they needed to start making a case for funding them in isolation. (CFAR was a spinoff from MIRI and 80K is part of the Centre for Effective Altruism that also houses Giving What We Can). Cognito is a standalone operation, so we have less of a time runway to experiment and demonstrate feasibility. However, as I wrote in the post, if Cognito is indeed promising and valuable, we hope to see our traffic increase while we're in maintenance mode, and we hope that people we've advised so far continue to benefit from what we've done. So we can revisit the case for Cognito a few years down the line with more evidence to decide if it's worthwhile.

Not sure how much relevant overlap there is. CM seems focused primarily on education, and on spreading relatively available but difficult to compile info to people who can derive value from it.

CFAR is largely focused on much less available content, and on the developing of new content, bith focused at more general needs.

Not so sure about 80K, although it is much less education focused. It also seems to have an additional purpose - providing support for people to do good, as well as do be more competent.

Neither appear to have a funding gap