I would err on the side of caution and say 2tps myself.
Maybe you should err on the side of doing the actual fucking math instead.
With 1mb blocks you can do roughly 7.76 tps when a transaction spends one output to two adresses: the payee, and the change address.
Well I don't think there's any need for that. ~7 is great on paper but in the real world it is nowhere near it, not even close. Many transactions have more than 1 input.
Isn't it interesting when D&T claims 2-4tps for 1MB your quiet as a church mouse, yet when I mention it you explode into an angry rage.
Still never mind, your going over to his thread now to tell him how wrong he is aren't you? Off you go...