Well...since I've already made an argument in favor of "progress" and declared my understanding for (what I assume are) Don's feelings on the matter, I'm going to go ahead and throw in an analysis of my real-world flashlight usage.
Months before I knew anything about Don and his Gizmos, I stumbled across the CPF Review of the Arc6 and Orcinus' photos of the same. I had previously decided not to get the Arc6 because I was confused about its modes of operation, and it looked rather frumpy in Arc's product photos. The review and owner photos changed my mind, and I decided to get one. When it arrived, I was shocked at how small and bright it was. I was temporarily disappointed by the fact that the PD switch didn't have latching capability, but after using it for a while I began to realize that this wasn't the obstacle I initially anticipated.
The first reason, which has nothing at all to do with the PD switch itself, is that I really do like twisties -- I fiddle with them constantly when given the opportunity -- and the PD design gave me an excuse to not give up having a twisty switch to play with, while still having the instant-accessibility that a pushbutton provides.
The second reason, which also has nothing to do with the PD switch, is that I prefer lights where the head screws into the battery tube, instead of onto it. I have absolutely no idea why, other than to speculate that it seems like a slightly stronger design, and/or because it makes the head seem smaller compared to the rest of the light. Then again, it could be nothing more than years of acclimation to the paradigm set by my old Arc AAA, I dunno.
The third reason is (finally) a functional one. While having a latching clicky switch is definitely advantageous from the perspective of 100% single-handed operation, that benchmark simply doesn't reflect my everyday usage. 95% of my actual EDC needs are covered by short bursts of light that last 5 seconds or less, so holding the button down for that span of time is insignificant. Also, as some have pointed out, being able to adjust the brightness of the light by adjusting the pressure on the button is more useful than mode-cycling during short-duration use, because cycling modes may actually take more time than the light will actually be pointed at the intended target. In those rare cases where I need sustained light and/or both hands free, I have yet to be so pressed for time that I couldn't afford the second it takes to twist the head to keep the light on.
So, while the McClicky+Aleph configuration may reflect an improvement according to one benchmark (i.e. 100% single-hand usage, time-to-activate notwithstanding), the PD configuration doesn't necessarily give up any functionality that real-world users would miss. This might explain why my Arc6 (now refitted with a proper titanium PD-Pak) is still the light I carry 3 days out of 5, and why there's a PD light on my hip 9 days out of 10.
Lastly, from a purely aesthetic standpoint: the PD configuration is what set McGizmos apart from other lights for at least a couple of years, and frankly, if functionality were all I cared about, I'd have bought a Surefire or two and been a couple thousand dollars richer at this point -- but it's not, and the uniqueness of the PD configuration contributed greatly to my perceived value of McGizmos as compared to other, more conventional lights. From a brand-identity perspective, it would be a waste to discard that uniqueness on account of its age, because in a lot of cases pure functionality simply isn't what buyers of "luxury" items care about; if it were, I don't think Swatch/Tag Heuer/Omega would still be building wind-up analog watches with mechanical escapements.
You know, that's actually a better comparison than I thought it would be, because it reflects how some people like being able to see and feel their tools doing their jobs, instead of trusting electronics to never fail. And even to the extent that the PD lights are electronic, they are FAR easier to service than Aleph-based lights are. I don't have the skills to do it all myself, but the switches can be cleaned, the emitters can be changed, and the drivers themselves can be modified. I can't do any of those with Aleph light engines; all I can do is buy a new part and throw away the old one.
I have to say, when it comes to pure aesthetics, while the Haiku is a master-work of machining and its interface is quite competent, the fact that it is unique in fewer ways than its predecessors makes it seem a bit more like an expensive also-ran than a truly unique offering. I seriously doubt I would ever sell it because of that, but it does see somewhat less pocket time as a result, which means ultimately it isn't fulfilling its intended purpose as well as its older siblings.
I know this doesn't reflect the "supply-chain problem" -- there's nothing I can do about that -- but I think it does address just about every reason why the "supply-chain problem" deserves a solution when it becomes practical to do so.