Some inferences of moderate validity about erosion prognosis might be drawn from examination of the vegetation in the high resolution photo provided above in post #32 by retired mech eng (https://www.metabunk.org/attachments/rx1hia2-jpg.24694/
). I hope that link is correct. I draw also from comments on geology by other astute and knowledgeable posters here, to wit, that the solid grayish rock is more competent than the brownish orangish rock and (certainly) soil. That fits my field experience and is backed up by plenty of science on the weathering continuum from competent bedrock, through "weathered rock" to "soil" which after all is just highly weathered rock with a higher level of biological content and activity than harder rock. It is never ideal to speculate without the benefit of a site visit, but I perceive areas of broken woodland canopy with shrubby, probably chaparral understory on grayish substrate, and areas where there is much more (bright green) grassy cover on browner/oranger substrate. Based on my field observation in analogous foothills sites, the former vegetation generally grows on more competent rock with a very thin layer of weathered rock/soil. (This is somewhat corroborated by the shallow depth of gullying of the gray rock, seen in photos in posts #30 and #31, above.) The latter grows on more weathered material with deeper soils; which is more susceptible to erosion.
The failure of the spillway occurs just about right where a band of the grassy/orangish soil abuts the spillway. There is the grayer, hypothetically more competent material downslope on the left side, and extending far upslope on the right side (viewed as in the photo; thus right bank and left bank of the spillway looking in the direction of flow). Where the browner, hypothetically weaker material intervenes, that's exactly where the failure is. There is clearly some very competent rock still sitting at the bottom of the waterfall, right in line with the chaparral/grayer material on the left side (in the photo). And, remarkably, a lot of the concrete (at least, the visible left wall) is still intact despite being subjected to a lot of energy impinging on exposed upper edges. I suggest that this means that the concrete that's placed right on the gray material, not on (heavens!) fill or the more weathered material, is not likely to erode much even under severe circumstances of extended periods of 100,000 cfs flow.
There is a narrow band of orange brown soil with darker green vegetation along the left side of the spillway; I speculate that that represents a body of backfill placed after excavating down to build the spillway. That's where the minor new erosion noted by Mick West is occurring, caused by water escaped at the sidewall breach above. It's either backfill or a tongue of that more weathered material; in either case, more susceptible to erosion than the gray stuff that's still happily bathing in the bottom of the waterfall.
Returning to the critical upper part of the spillway, above the breach, the woodland/chaparral on grayer material seems to run along the whole of the right side. There is slightly higher headcutting on the left side than the right, again supporting the statement that the woodland/chaparral/gray substrate is more resistant. It is not reassuring that the headcut scarp on that right side is brown/orange material, but the exposed rock peeping out through the vegetation above is gray, and hopefully that means getting more competent as one moves uphill. In any case, I think the pattern of vegetation and other clues suggests that the hard gray rock certainly extends (at the surface, overlain immediately by the concrete) partially underneath the intact upper spillway. Probably the construction documentation does not even show precisely its extent there, but the fact that the headcutting did not propagate much uphill from the original failure of the concrete suggests that it probably extends right across. Given how well pieces of the concrete in the lowest gray-rock area have remained attached, it's plausible to infer that the upper spillway is also well attached, in close contact, to competent rock and will probably not fail and headcut all the way up, even if fractures develop or additional pieces of the slab are lost.