As someone with plenty of experience noted recently that the earth (dirt) moves around for all sorts of reasons (water saturation to bone dry cycling) over the years. The spillway has been there for a while and its subsoil foundation apparently developed some problems that made some of the weighted slabs drop into a void.
It would be interesting to know what the downstream flood stage water flow capacity (including the irrigation diversion channels) is. I'll bet that it is less than 100k cfs.
I saw a clip from a helicopter view this morning. Not far downstream is another dam with gates, which diverts water to the west, towards what appears to be a giant man-made holding pond. I’m sure that thing is getting full. And the diversion dam...looked about to get over-topped. That was this morning - who knows how bad that got during the night.
Then they showed the town of Oroville. The water still had around 15 feet before if took out the bridges, but the river was either way up on its banks, or even out of it. Who knows what happens when the land flattens out further downstream.
I’ve designed a few (much, nuch smaller) dams, and know a little about hydrology. The thing that struck me was the massive water surface ‘jump’ immediately downstream of where the spillway structure met the flume. I think this is what ripped apart the east sidewall of the flume and caused the water to rip at the flume from the side - and that is what caused the void...but the initial failure was the sidewall falling apart. The stuff people are talking about with rapid draw down or fill up usually deals with the water side of the reservoir - not the dry side of the dam.
In 1961 (when they designed the dam), predicting that jump would have been a paper exercise...and you could probably get whatever answer you wanted. I’m amazed that sometime in the last 50 years, the spillway flume was never re-modeled using computers, etc.