There are some limiting principles... the lower decision only covered books that were "in print" in eBook form... but the rationale here is quite broad and would easily stretch beyond these specifics. (There's a small amount of analysis related to whether the digitization involved in CDL is "transformative" that rests on official publisher eBooks being available, but there's a strong overall impression that the decision would come out the same way for things not already available digitally.)
In a more reasonable world we could imagine Congress might pass a law authorizing actual one-copy-per CDL by non-profit libraries. But nobody's going to hold their breath for that.
I've thought since the beginning of this saga that a change in statute would pretty much be needed for CDL or something similar. The idea being to craft something that extends the philosophy or idea of libraries in the face of an increasingly digital world where doing much of anything requires a copy, things are licensed rather than sold, and the first sale doctrine has little application, but I agree with you that such an idea has dim prospects.
But looking at it, maybe this was the way the original suit was set up necessarily, and the lower court decision? OK.
IMO, properly executed CDL had/has a good chance of succeeding.
This is actually a pretty significant limitation, because so much of what was practically available as CDL was actually out-of-print books that the publishers never bothered to make available for eBooks licensing. It's at least reasonable to expect that the fair-use analysis might tilt the other way for such books - the use is a bit more "transformative" because at least it technically contributes something that the publisher didn't, and the potential of market harm wrt. the copywritten work becomes a lot more speculative.