I feel like it's time for AVRs to come with a full blown OS so that 3rd party providers and independent developers can create apps for it. Such as built in media players, DSP plugins, or streaming service app.
It is called a HTPC.
It is not whether you have a full-blown OS but whether you have an open architecture. Even then, it is not clear you need it to be integrated with the guts of an AVR.
The thing you cannot have in a DIY HTPC is multiple input connectivity. That would require transceiver cards that would work in the OS you have whether it is Windows or Android or Linux or some proprietary OS. Such cards do not exist. Because there is no business case for them. So, just putting Android or Linux is not going to solve this.
Once you make the OS participate more in the low-level functionality at kernel level (which are not required to support the needs you have mentioned above), then you get into real-time constraints and need an RTOS with modular design. NAD and Trinnov have developed such OSes with modularity but they are not open architectures. If they opened their architectures to third-parties, then they would be what you are asking but the business case is difficult to justify such a thing.
If you have a proprietary API, you won't have third-parties developing for it because then they are tied to the market for that one device. If you assume there is a common standard open API across multiple devices, then it requires all of the manufacturers to agree to it. There is no benefit for them to get into this. Nor does it make business sense for all the support headache they will have with users trying to integrate all kinds of third party software. The benefits of opening the architecture must far outweigh the cost of doing so. This is not the case for niche domains like audio/video.
In the purely software world, such things can be done as open-source efforts but open-source has limitations when hardware integration is needed. You have limited success with platforms like Arduino but it only caters to a fringe and niche nerd market. In any case, the need to have licensed connectors and certified DRM handling would keep it out of open source projects.
An open architecture only makes sense if one of them vendors get really big like Apple or Microsoft and can justify a whole ecosystem on their own. Audio is a vertical niche market compared to a general purpose PC, so I don't see this happening in practice. There is no technical reason for not being able to do this.
Smart TVs are a limited example of such open but proprietary architectures and it is only possible when you have the kind of volume Samsung or Sony have to attract third parties but even then, they have to coax/bribe third parties to come and develop. But what they have done to make it easier is to use one of the standard OS as the "application server" to host the apps. But TVs have limited capabilities compared to an AVR.
Of course, that would make it harder for manufacturers to come up with an upgrade for the next year. I guess this will have to come from some outside competitor that is new to the market.
No. The "outside competitor" has far too big a moat to cross in this relatively niche world and even if they managed to cross it, would find themselves in the same business model quandary as the rest. The value to them comes from the integration not open architectures.
Problem with AVRs is also that they have very little space to accommodate a full-blown "application server" inside them. The more "full-blown" it is, more space and power/heating issues to solve.
But I do foresee a possible move towards a limited "application server" like the Smart TVs for content streaming in future AVR generations but not for hosting third-party content processing that requires hardware.
Sort of like Sonos and Denon converging from different directions.