https://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:77219071-1

and

https://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:77219073-1

TLDR, I am in favor of treating both species as species and will probably go ahead and draft the swap.

Long story: I consider both questions unresolved but think the swap will be moderately useful and, most importantly, won't do any harm.

That said, there is strong evidence that E. spurca and E. theriaca occur close to each other and occasionally sympatrically and I know of no credible intermediates (the plants the look intermediate are clearly drought stressed). The problem is that the characters that distinguish them are known to be unreliable across the section (though stipitate vs. sessile glands works in almost every case if you pay attention to regional variation). There are trends in seed differences as well, but I never got a chance to evaluate the NM material while I was working on the group and, as a consequence, was never able to figure it out. E. therica and E. spurca are extremely easy to separate in the field compared to E. theriaca vs. E. simulans or any of them from a sp. nov. I have yet to describe. Except for the seeds, you wouldn't guess they are sister species. Yang and Berry (2011) did sequence the two and found them to be very closely related. As such, it's possible that either occasional introgression keeps them from becoming totally distinct or they are only recently diverged.

For E. triligulata, I am actually investigating the problem for my PhD and will probably have data this year or next. Currently, I think it may be more closely related to either E. scopulorum (compare with the only known photo of E. scopulorum var. nuda on iNaturalist) or E. fruticulosa. That said, E. chaetocalyx can adopt some pretty large rootstocks so it's hard to rule out the possibility that it originated from a population of E. chaetocalyx.

As for Powell's treatment, I reviewed it prior to publication and think it's very good. The decisions that led Turner to consider them separate species before him, not so good. Though, Turner's rational for recombining E. triligulata with E. chaetocalyx was even worse, so I suppose it balances out. All that set aside, the species are distinct enough for me to back the switch until new evidence comes to light.