process contributors could be more efficient with API calls

We might be able to squeeze a lot more performance from our api calls in the process_contributors task

Sonnet 4.6 has identified that we are not being efficient for large repos:

https://github.com/chaoss/CollectOSS/blob/96adf3a4d68725db21622673ee6613693c0f5ace/collectoss/tasks/github/contributors.py#L51-L62

> loads up to 500 contributors missing cntrb_created_at, then hits GET /users/{login} one at a time. This is essentially the same pattern as the comment collection win — it could be replaced with batched GraphQL user lookups (up to ~50 logins per query with aliases).

this batching behavior is already present here:

https://github.com/chaoss/CollectOSS/blob/96adf3a4d68725db21622673ee6613693c0f5ace/collectoss/tasks/frontend.py#L207-L212

(note: this issue is very similar to several others. be careful to make sure you are talking about the same issue in the comments)