This whole argument makes no sense. No one says you can only review branches when they're "done", many open source projects use RFCs, then WIP branches etc, no

This whole argument makes no sense. No one says you can only review branches when they're "done", many open source projects use RFCs, then WIP branches etc, no problem. The whole issue seems self-inflicted and the problem goes away when you remove the initial wrong assumptions.