You're not sure? Seriously? Surely you must jest. Nintendo has like a gazillion users historically for gaming and Apple has like 1000 people that want to game on it. I'm one of those people (don't really want to own two computers anymore and a console is not a direct replacement for PC style gaming), but Apple doesn't CARE about 1000 people. Of course if it had better gaming capabilities maybe other people would consider a Mac for gaming in the future particularly if it can also run Windows 10 in the mean time. Is it really so hard to offer a REAL GPU on a desktop Mac, for example? Surely Thunderbolt 3 could handle an external hub with a real graphics card. Certainly the richest tech company in the world can afford to throw together a product or two and see how the market responds? No no no. We have to make more phones instead.
Oh for the days when Mac Vs. PC commercials were actually funny instead of the Apple phONE. (no longer the iPhone in V8. Have to follow the new convention. The "i" has to go! Apple phONE running phONEy OS).
The big difference is that Blizzard has a decades-old experience coding games for Mac (probably as much as for Windows), while they have none coding for the Switch. (Noone outside of Nintendo has any experience coding for it). My general philosophy about business moves and decision-making is this: if a decision doesn't make sense on the surface, then follow the money. That philosophy has yet to steer me wrong, so you may find it useful, too.
It's too easy to go to the comparative market sizes, because we all know that Linux and MacOS come out on the short stick on that one. Although, SteamOS/Linux focus by Valve demonstrates that disruptive things can still happen in the gaming market. We don't know what Apple's long-term strategy with regard to gaming is (other than iOS is higher up on that ladder, but, like it or not, Metal is still in its infancy as far as graphics APIs go. In DirectX terms, we're still a year away from Windows 98. Getting too upset because Metal hasn't yet achieved feature-parity with more mature APIs is a waste of time and energy, and I don't have enough of either of those to waste on it.