Ansible galaxy concept is great as it supposed to promote playbook reuse.
Still because it lacks any serious filtering or curating methods it became a victim of its own success: there are too many options to pick from and most of them (if not all) being incomplete, obsolete or without a community behind them.
Ansible galaxy reminds me a little bit about Jenkins plugins, another place where too many options is worse than having less.
Do you know a way to deal with this problem? How we should approach this issue.
I would love to contribute to incomplete roles, but with the current state of affairs it seems impossible to know which one to worth any attention.
There are a couple of different approaches to this. https://galaxy.ansible.com/explore/ shows you a couple of different criteria by which to browse available roles. "Most Downloaded" are obviously the most popular, which means they are either the most complete and/or are being constantly maintained. Similarly with "Most Starred" list.
You can also browse roles from the most prolific role authors. Generally those people spend a lot of time writing their roles, so they tend to be up-to-date and follow good practices. If that's not enough, you can simply look at the last update timestamp of a role you are interested in. If it's been a couple of months since the last update, perhaps it's been abandoned.
For common tasks there will be a lot of similar roles, so look at the most popular ones. For more esoteric stuff your options would be more limited and easier to narrow down good candidates.