Hi
Before all, thanks for the wiki page tracking the payjoin adoption, it is a good idea.
-----
Even when there is a reasonable economical incentive to use segwit transactions to save fees a big percentage of the transactions are not using segwit yet. In the case of payjoins the economic incentives are not so big while the privacy benefits are not so clear for the payer as they are for the global transactions graph as a whole. This means that payjoins requires some level of altruist attitude from the payers. The payjoins UX is also not good because I think most users are not familiar with bip21 uris (users still request support because they pay a bech32 address in an exchange and the exchange tells them that's not a valid bitcoin address). All this is relative and subjective but in general terms I would say it is more or less true for many people.
Anyway, imagine wallets' developers agree on making payjoins payment by default because it is the right thing to do (fight against surveillance to spy on
bitcoin users and improve bitcoin's fungibility). In that case it should be completely transparent to the users and at not cost, it shouldn't require the user to do anything different, it shouldn't be noticeable slower, etc. In fact, users should have to know they are payjoining at all.
The only way I see to achieve something like that is by moving to schemes where wallets can communicate and interact. I should be able to know something about you that allows me to select your name from my contact list and select "Pay to Chris" and if my wallet knows how to find yours then it can request a new address and pays, or generate a new one for you (probably using a output descriptor you created to share with me).