The recently stated goal is to decay to 1 infection per million residents so that robust contact tracing works. Back of the envelope math puts that between 6 and 18 months of continued progress of the health order to get it to this level.

(60 or so in Santa Clara County ICU == ~6000 infections. ICU occupancy decaying by 40% per month; 0.6^15 months * 6000 = ~3 )

If you look at hospital admissions, instead, it's a bit less than a 40% decay per week. So we get something more like 5-6 months.

I don't see either of these as plausible. And even if we have lowered Rt to this point of decay for most subpopulations, I'm better there's a few where it's higher and that will eventually shallow out the decline. Not to mention the risks from areas with lower controls.

The IMHE model is very influential, but it has been forecasting an unrealistically quick decline in cases that has encouraged public health officials to seriously consider trying to get to this minimal level of cases.