Following this thread closely. We are about to face the exact same decision and the replies here are gold.

Tried this on a staging copy first and it behaved exactly as described. Confident enough to roll it to production now.

Great point. I had not considered the angle you raised in the second paragraph — it reframes the whole thing for me.

We ran into the same situation last quarter. What worked for us was starting small and only adding complexity when a real need showed up.

For what it is worth, the same approach works with the PMPro adapter — the membership side is identical from here.