One thing to consider is that PDS now seems to allow ~3 years for a major new GSG. That means starting developing for release in 2027, which gives HoI4 over a decade. I think that's reasonable.
But I wouldn't want to see HoI5 in the near future for four reasons, moving from the more personal to the general.
Firstly, PDX have said that they are dropping support for Linux in future games. That puts me off buying HoI5 at all.
Secondly, I have a nasty feeling that HoI5 would move to 3D character portraits, as CK3 and V3 have done. Personally that's annoying because I built my current PC to meet the CK3/V3 minimum graphical requirements, and I would resent having to buy another GPU to play a GSG which doesn't really need top-of-the-line graphics. And that seems like a strategic mistake to me, since every penny or cent that I spend on GPUs is not being spend on PDX DLC or subscriptions (my gaming budget is pretty limited).
But aside from personal gripes, I also think that 3D portraits are a significant impediment to modding. Modders like those who make EPE for CK3 do amazing work, but 3D modding is always going to be a niche activity. For one thing, the PDS-recommended workflow uses Maya; in my country a Maya subscription costs 10% of the monthly median wage, so it turns character modding into a hobby for the rich. And 3D modelling is hard work (I have learned enough to make 3D models for Simutrans and it's one of the hardest things I've ever done, even compared to e.g. learning Mandarin). This is not a huge problem for CK3, because we don't know what a lot of the historical characters looked like anyway. V3 does have some mods adding historical characters, but the biggest (ECCHI) hasn't been updated since 2022. Historical characters are much more important in HoI4 (I would guess that the generic portraits are only used by players in a minority of playthroughs). With 2D portraits, modders often just use a photo rather than drawing an portrait, and there are now LoRAs (AI models) that allow you to re-create photos in the HoI4 portrait style. They're not perfect, but they're getting better fast. If HoI5 uses 3D character models, then I think a lot of modders will be put off trying to include historical characters in their mods.
Thirdly, there are still several parts of the world that were very important theatres in the Second World War but don't have official content. The ones that particularly stand out to me are the Middle East and South East Asia. And I think that each of those areas could easily justify two DLC. Assuming 3-5 tags (or groups of tags) per patch cycle, the former could be split into the Near East (Egypt—Levant—Maghreb) vs the Middle East (Afghanistan—Iraq—Iran), while the latter cannot be properly handled without distinguishing maritime South East (Dutch East Indies—Malaya & Borneo—Philippines) from continental South East Asia (Burma—Indochina—Thailand).
Fourthly, the game still has many long-standing bugs that have not been fixed, particularly in naval warfare. There are too many National Focuses, Spirits, Advisors, and Doctrines that promise that have an effect on combat but actually do nothing. Having paid for a working product, I really want to see this fixed. I was impressed by the way that Arheo got the bug count down to a really low level in the final patch of Imperator:Rome, and I hope he is aiming to do the same again, even though HoI4 is now a considerably more complex codebase.
