My experience is exactly the opposite: that the abstract formula should be displayed as soon as possible, then justified, then exercised. My philosophy on this has been developed by the idea that I want the formula in the student's visual field for as much time as possible, in the hopes that it will sink in mentally. Note that this is sympatico with the requirement in many locations that the teacher clearly state the goal of the lesson at the start of a meeting. Plus: The cycle of theorem-proof is simply traditional mathematical presentation and writing style (for exactly this reason, I think), and students should get to experience and expect that style of presentation as a "real" math class.
Personally, I always get weirded out when I see instructors doing the opposite. They seem to take most of the presentation time doing these warm-up exercises, and wind up squeezing the ultimate goal in the last few minutes of class (and not actually exercising the formula itself). As both a teacher and a student myself, I feel that we all get confused about what the "point" of the lesson is, what the real take-away skill is, and how it will be assessed in the future.
I understand that many of us wish that we could lead all of our courses and students in "discovery" style lessons where they take personal ownership for all the new material. But as teachers this is simply infeasible granted the limited time in class; and particularly so from my perspective in the college classroom (even if much of my career is spent remediating topics from middle school).
(I was reading material by Hung-Hsi Wu recently and he does this always-concrete-warm-up cycle, and while his material is otherwise excellent, I find this distracting and it forces me to flip back-and-forth a lot to uncover the "real" proofs.)