Some organizations maintain lists of people who have either applied to roles there or who might be good candidates for future roles. They often share these lists around when other orgs in their space are hiring. (I only know directly of CEA's list, but when I was recently looking to hire a content person, I got ideas for candidates from staff at many other orgs, often with quick turnaround that made me think they had those names on hand already.)

Typing at the speed of thought, not very confident in any of the below:

This project seems reasonable for someone to try. Obstacles that might come up (none of which seem like compelling reasons not to try):

1. The usual "this will take a long time" obstacle that goes with any big technical project. Building this well will take hundreds of hours of someone's time, and job seekers will spend hundreds of hours of additional time adding their information to the system (if it ends up large enough to be broadly useful). Maybe the system will help to save even more time for hiring managers, but it seems like it would have to be quite successful to do so.
2. The categories you've listed don't track the most important thing — someone's actual skill. When candidates are referred from one org to another, it's often in the context of "this person applied for our role and almost made it" or "this person we worked with was impressive, and you should hire them because we don't have a role for them". That's a lot more evidence for reaching out to them than a tag on their resume.
3. This doesn't seem like it actually makes it easier to get hired. It could make hiring managers' jobs easier (which is valuable), but if anything, it seems like this would increase the # of applicants per position (if orgs find it much easier to source lots of candidates). This is good for the orgs (and the world), but makes each application process more challenging for each applicant.
   1. Put another way, someone whose struggle was "I applied to ten jobs and got no interviews" would now experience "I added my resume to the system and no one ever reached out", which seems even more frustrating.
   2. On the other hand, someone who applied to four jobs and got four interviews might get a lot more reach-outs this way (from jobs they hadn't considered or known about), so this platform's core function might just be "matching the best candidates to jobs more easily", which is a different sort of thing but still good.

I also wonder if there's a way to do this through a platform like LinkedIn that people already update regularly. I'm not familiar with that platform's search capabilities, but if there were some way for people to join an "EA group" on LinkedIn, maybe hiring managers could just filter for people in that group (and then take advantage of all the other LinkedIn filters)?

Another way to get a bunch of info added would be for several orgs to agree to handle their hiring through it — that is, rather than having their own resume collection systems, they'd ask candidates to add their information to the centralized system.

There is the EA Hub profiles directory, where you can search for people by location, cause area, expertise, and whether they're open to job offers

I tried to do something like you're suggesting with this longtermist census.  

There a trade off in how public you make the results.  The more public the information is, the less information people are willing to share.  I wanted to ask questions getting at "how much do you want to change job right now" and so decided not to make the results fully public.  

Hey, that's a pretty interesting concept you've got there! Creating a "resume book" for the EA community could be a fantastic way to streamline the hiring process and connect potential candidates with organizations more effectively. Your ideas for a minimum viable product and an advanced version show a lot of potential, especially the recommender system and the integration of EA job listings.