I'm just saying, what's common sense insofar as their multiplatform support, realistically speaking? The average person would assume it extends to PC and limiting any other console support to just certain sports games (like MLB, which wasn't up to Sony but rather the MLB League) and specific live-service titles like Bungie's output in things like Destiny 2.
There is no logical reason to assume, going off their financials from PS division, why they would need to potentially jeopardize their console brand image and marquee status of defining IP, let alone taking on the burden of splitting dev resources among more hardware profiles (meaning less optimizations in similar time frames than if they just focused on their own hardware), by bringing anything that isn't a specific sports title, or certain live-service GaaS titles, to non-PlayStation consoles. Especially considering that in such a case we are at that point talking about PS Studios games, when we do not see the equivalent from XGS titles going to platforms other than Xbox & PC, or certain Nintendo games going anywhere else but Nintendo platforms.
There is no financial situation where Sony would be forced to compromise their console's brand image and selling power by doing such a thing, no wider long-term industry-related situation either (before you say "chip shortages!", they can pay for priority and figure out new product ideas that still keep vertical integration within their own specific hardware line, increasing revenue and profit potentials and retaining independence). So it's rather easy to figure "how much more" they will go multiplatform, you only need to look at the actual trends on their end and be honest about their market position and division results. Which, when you do so, should be easy enough to realize that multiplatform support will very likely remain PS & PC, PC Day 1 only for certain live-service GaaS titles, and Xbox (maybe Nintendo) support only for select sports games and GaaS titles. IF Sony happen to purchase another major publisher who has key IP established on, say, Nintendo, then their multiplatform support would probably extend to retaining those games on such a platform, perhaps with some timing of release changes depending on any internal developments.
25 million people would not be paying for Gold, however. Your calculations are wrong because you're assuming every GP subscriber has converted from Gold, but if that's the case then XBLG subs would be near zero if you take typical install base > subscriber base ratios into account which we can see from PlayStation and Switch.
Now, we don't know what rate of XBLG subs have converted to GPU; it's probably more than 50% but it's definitely not near 100%. Which means you have some amounts of people on GP who are paying month-to-month, but likely aren't paying for the entire year. Then you have others who might be on a free trail and don't sub afterwards, others still who are using MS Reward points to more or less get months for free, etc. Also, in certain territories GP/GPU subscription rates are a lot lower, India for example is one such place.
So no I'm still rather confident in those numbers from the other thread; if those were based on misinformation then you're basically saying not just the article, but several of the websites used that track market data, and even Sony and Nintendo's own fiscal results WRT services, are also misinformation. I did say the calculations weren't perfect because it's extremely difficult to line all that data up in a simple timeline/timeframe, some results have to be extrapolated from guessimates that are reasonable enough etc. But realistically I don't think those calculations in the other thread (particularly the last ones we came up to) are that far off from where GP annual revenue actually sits.