This is an interesting perspective. I've recently been playing with sequent calculus, and I'm writing everything upside-down because I'm starting with the statement to prove and applying the derivations mechanically to produce the proof tree.

But hearing how you approach proofs makes the standard directionality make much more sense.