> especially because a lot of features already in bronze are not things we actually use or need
This has been my org's issue as well. We use dedicated, external products for project management, CI, binary asset storage, container registry, wiki, etc etc. I get that a one-stop shop might be really valuable for a small org with basic needs that align well to what GitLab supplies, but I don't know if it scales up all that well. I would worry about getting partly migrated and then discovering some functionality gap, and maybe finding out that the GitLab team that built that feature no longer exists and there was basically no one maintaining it any more.
Spreading themselves so thin does have a real cost— here's an example of recently-resolved ticket for something that I would consider core "repository" functionality (LFS content inclusion in the tarball) but that sat unfixed and unnoted-in-docs for almost four years: https://gitlab.com/gitlab-org/gitlab/-/issues/15079
This has been my org's issue as well. We use dedicated, external products for project management, CI, binary asset storage, container registry, wiki, etc etc. I get that a one-stop shop might be really valuable for a small org with basic needs that align well to what GitLab supplies, but I don't know if it scales up all that well. I would worry about getting partly migrated and then discovering some functionality gap, and maybe finding out that the GitLab team that built that feature no longer exists and there was basically no one maintaining it any more.
Spreading themselves so thin does have a real cost— here's an example of recently-resolved ticket for something that I would consider core "repository" functionality (LFS content inclusion in the tarball) but that sat unfixed and unnoted-in-docs for almost four years: https://gitlab.com/gitlab-org/gitlab/-/issues/15079