I found that it is possible to get it to work in a non-obvious way. I saw a post somewhere in my searches from someone who mentioned using the “project ID” - the long number you can find in the URL when looking at a project - in the zap to correctly connect with that project. Each time it takes me some fiddling, but it seems to work out using that.
So that was a “hack” way to make it work for us, but I would hope for better than this from Asana.
My solution uses a google sheet as an intermediary to log when the last sync occurred. I use a filter step to stop any sync that tries to run within a few minutes of the last full sync, and a ‘delay queue’ task before the filter to keep incoming changes from asana delayed and only entering the zap one at a time.
It works fine, but it’s a 3-action workaround that shouldn’t have to exist.