I'm not convinced anything has gone terribly wrong with the forecasts for this event. I think SPC has just overplayed the daytime threat, and possibly the entire day, too (based on how far west the 30% tor area was). A lot of the CAMs did not exhibit a strong signal of widespread rotating storms; there was a lot of variability among them and some were more bullish, with others being more bearish.
Take this 12-h forecast sounding from the 12Z 3 km NAM for OUN:
It's got the sharp capping inversion and everything. Looks pretty comparable to the observed 00Z sounding...maybe even
too strong with the cap. But it also only predicted one real string of UH with this event in OK. Sure, it was very high in magnitude. But keep in mind that's also just 2-5 km UH. That says nothing about tornadoes...just mesocyclones. I'm sure we've had some long track rotating storms today, but the poor shear in the same layer (2-5-km) exhibited in multiple 88D VADs suggest the profile just wasn't great for sustained supercells (see below):
Yeah, shear looks great from 0-2 km, which can get you tornadoes provided the environment doesn't get spoiled by cold pools or other interference caused by crappy supercell dynamics. And I think that's what we have.
Also, it remains the case that the synoptic scale forcing for this event was always displaced to the north, in KS and points northward. I am a bit more mystified at the lack of storm coverage up north, though.