This updates to the latest available (and satisfiable) version of every package you have installed in your environment. This is the best way to ensure you are running with the latest bug-fixes (but creating new, small envs for each workflow is better than having a load of packages in your root/base environment)
> conda update anaconda
You are updating a single meta-package here to the latest version of it. This meta-package is released roughly 4 times a year and specifies/pins the exact versions of all the packages in the Anaconda Distribution installer which will have released at the same time. The latest anaconda meta-package is from October last year so it does not identify any software builds newer than that. This means it will downgrade more recent packages versions.
So conda update anaconda is useful in 2 situations:
1. When a new AD release is made and you are running with an old one and want to update that to get all the versions we've tested and have determine to work well together.
2. If things break with new package updates and you want try to get back to last set of versions we've tested and have determine to work well together.
But my recommendation is to just use an env for each workflow that contains only the software you need for that workflow and to keep it updated with conda update --all.