From my perspective can't say I've seen all that much in the way of improvements, in general use, on my three Windows Phone's for like 1.5 years (probably longer). Still same old "resuming", slow to load apps / perform functions many a times, non-fluid feel, inconsistencies still abound, and so on. IMO these updates have done little so to me they don't mean a whole heck of a lot in the grand scheme of things.
Until Andromeda becomes mainstream (and MS doesn't kill it beforehand) not sure we can say it is truly alive either... can we?
Andromeda is both over and under emphasized. People don't really understand it's place in the scheme of things. It's a sort of beta hardware, like the current HoloLens. It's also one of the first windows core releases towards a unified crossplatform OS. So it doesn't have to take anything by storm.
Probably by the time MSFT comes round to make andromeda 2, Samsung will have perfected the janky inbetween hardware, that will serve us until graphene flexible screens become a thing (from what I can tell, what they appear to be working on for the x, is a combination of small amounts of flexible graphene, with a metal watchstrap style series of fixed panels, a little like the surface book hinge). And MSFT will have its foot in the door as far as a software platform to use for it. But both of those devices will be expensive I suspect the first x will be like 3000 usd.
But more importantly, as one of the first versions of windows core, it provides one of the several puzzle peices that will eventually unite to make a single OS to run on every platform. And that, will give windows the potential to adapt, as the market slowly shifts, later on, towards AR, VR, flexible screens and so on.
I can see no reason why, given the stated goal of MSFT since windows 10 released, of creating a unified platform - a mantra they still repeat, why they would not release a windows core version for every FF of hardware. Essentially, for their investment so far, all these years, to mean anything, they basically have to. Even if the whole project is a flop, or only a few thousand buy it like the HoloLens, they need that opening, to expand on later.
It's not like the candy bar smartphone, which is a limited lifetime technology and has already reached peak saturation and is slowing - where it simply makes sense to focus on the next battle instead. These are battles that won't be fought for years to come, in any proper sense.