I know I've never attained the kind of football success that will cause a major research university to overlook my moral shortcomings, but is it me or did he just explain what a spread option offense is, and not really what a RPO offense is (like Moorhead at MSU or what the Eagles have been doing)?
I was under the impression that what sets those offenses apart is that they're reading the reactions of a specific linebacker or safety quickly (whereas all of those examples Urban is reading a defensive end), and if they commit to the run, you yank the ball out and throw it to where there's now a deficiency in coverage. The read has to be fast, and the ball has to come out quickly, or else you have linemen running too far downfield on passing plays.
Whereas Urban's spread identifies a defensive end to isolate, the RPO can create mismatches across the field (both horizontally and vertically) using formations, and then target a specific mismatch to make sure that, if the read by the qb is correct after the ball is snapped, what the defense does is always the wrong answer. While Urban's defenses are designed to spread defenses out horizontally, the early RPO offenses solved the problem of not having a running qb by adding a quick throw out to a WR (like what they did with Peterman at Pitt), but have since evolved to take the same basic idea and attack vertically, which is what Osovet was known for.
The hard part, however, is finding a Carson Wentz or a Trace McSorley who can assess a defense and make a read really quickly as the play is happening, and then fire the ball out quickly and accurately anywhere on the field, and not just to the flats like Urban described.
Anyways, just felt like what Urban described is one or two iterations behind what some of the more create RPO offenses are doing now.