Maintaining perpetutal support and dual firmware builds is not anything like a 'minimal effort' when every device manufacturer and ecosystem manufacturer has to do it. Perpetually. Because collapse could strike any time. I've done a lot, and I many years and years, on client/server (device/server) work and it's never minimal effort when you have to support just one secure, robust implementation let alone a readily available shadow option for something else. It is literally two different builds both of which need to be tested every time a change is made on ether side. This is time consuming and expensive. Even huge companies like Nintendo can barely cope with one decent online system, now they need a second?
Literally my last post started with the words "Nothing is an unsolvable problem". Technically. I don't know who you are quoting with 'This can't happen' because I've never used those words. But your minimal effort is anything but minimal effort, its metric fuck ton of running costs and client/server code testing forever. Technical and massive cost hurdle aside, where I started out is that this simply won't appeal to digital content providers from a commercial standpoint because fundamentally it requires competitors to set aside their competitive nature while continuing to operate in what is a very competitive market. This is a high effort, high cost solution to something that does not benefit the people paying the bill.