Yes, I have some experience with UWA development, although even in the space of app development, I have way more experience with iOS & Android development, for obvious reasons.
The moment you have a plan to add a platform or a new distribution model to your game, your cost in terms of funding & effort will go up. That doesn't mean it makes it unfeasible, just that those two costs goes up, and that we'd only choose to do this if the benefit of this wider platform release compensated us enough for it. It would only be prohibitively expensive if UWA development wasn't bringing us enough of an ROI necessary to do so. Making our game .exe gets us on steam & on every major console. If UWA development just supplants Xbox & PC, two platforms my .exe already runs on & makes money from, why would I focus on UWA development?
Ideally, if we're talking about creating a platform that software developers target, then we need to take a step back from console development, and look at this in the broader view of targeted OS development. Right now, if you're making new software, you're focusing on app development, and you're targeting iOS & Android first & foremost, with iOS probably being your first choice. UWA's just haven't taken off yet as a tertiary platform in that sense. I don't approach developing software for Windows as a UWA, because I can still reach my audience through Windows regardless despite it not being a UWA. I do think that, ideally, MS would want software developers (not just game developers, mind you) to approach developing software on Windows as UWA, but they also know they can't force us to do it. Their current approach is to try to make it as seamless as possible, and eliminate as much of the effort cost as they can that it would normally require us to put in. I appreciate their efforts, but they still have a looooooooong way to go.
The reason why its different is because game devs write our games, particularly console games, to have our code bases specialized towards the machine they are running on. I was discussing with another user a few of the possibilities that we might see in order to get such a scenario, but none match the current discussion we're seeing arise from a mid-step iterated console. Even when comparing it to PC games development, which has its own unique & wild set of problems that comes with the open-ended platform that it is, it's still very different in terms of execution. So, unless the console's APU & OS is identical to the weaker Xbox, then I have to build a separate code base to have my game run natively on the machine. Which would then require it's own amount of debugging & QA. Who is eating all of the financial & effort cost for that? Me, the developer. And since its just another Xbox, all this results in me selling into the same market ecosystem i'm already selling to; so its increased cost for no extra gain.