Pentax and architecture
Ricoh has announced the end of the K series of cameras, to no one's surprise. Here is what I think is going on. First, Ricoh Imaging has been booming and the unit is now debt free. This came on the back of the Ricoh-branded GR line. They are now as well positioned to now make investments and pivot as they have been in years (or exit cameras altogether). If Ricoh doesn't sell the unit, I think they have learned that niches are probably all that is available to them, but that niches can be viable.
And then there's the supply chain: Pentax uses Socionext processors, and Socionext is itself focusing on AI and has developing other business lines (e.g. automotive). I seriously doubt that Pentax is calling the tune with Socionext. We have seen camera manufacturers issue new models on the basis of component changes in the past, and Ricoh themselves have blamed supplier issues for ending some products in the last year or so if I recall correctly. And the K-series bodies appear to have a much more complicated architecture than the GR series, with several chips compared to just one. I would guess this has happened to iteratively work around the limitations of the mirror-based approach as they attempt to solve for uses like video for which the platform was never built. I don't know where Ricoh is sourcing pentaprisms and mirrors these days, but those providers will surely be losing interest.
Ricoh mentioned Pentax's K mount. at this point I think the glut of used lenses is now working against Ricoh; there are sellers on Yahoo Auctions here in Japan looking for as little as ¥15,000 for a wide range of lenses - even some decent constant-aperture AF lenses are very cheap.
So it seems the market is narrowing, supply chain dependence is throttling profits, and their own legacy glass is hampering sales of new lenses. They have to go where they can find a sustainable position. Of course, the market is insanely tight: a Lumix S5ii (a "full-frame" mirrorless) body with a well-regarded kit lens is $2,700. Competing in the mirrorless already space fiercely contested by four three major players plus a low-cost contender with limitless backing would be insane.
I suspect the point of moving to a new line of bodies is that the re-architecture that gets them off their complex past. I am guessing that whatever they build, it will be much simpler: a replacement for the mirror & pentaprism, a much simpler chip set, a cheaper body (subtracting metal, adding plastic), and decent autofocus. I can't wait!