How do you implement that, non-ambiguously, in an intuitive way?
The second link has a bit about that. As it turns out, you really don't want real-world physics. In the case of SMB, much simpler models (see the fourth link), using nothing more than simple arithmetic (in line with TFA), are not only effective, but much simpler to implement on old hardware like the 6502. The iterative process of adding effects and tweaking values until the controls "feel" right. (I can't find the interview i saw earlier, but this should give you a sense of that iterative process, from the developers themselves.)
Sure, anything can be accomplished eventually through brute force and pure persistence, but the programmer with math and physics knowledge is going to do it better and faster
No one is disputing that such knowledge is useful. The question here is whether it's essential.
For games, it's pretty obvious that it's unnecessary, as you admit here. Though I'd like to add, from some of the discussions I ran across, attempting to model real-world physics in games can actually introduce problems that simpler approaches avoid. Again, see the second link. There were a few discussions I ran across yesterday with a quick search that go in to this as well. I can dig them up if you're interested though it shouldn't take you more than a few minutes to find them on your own.
and likely more elegant, because it's hard to find something more elegant than the laws of nature.
I keep going back to that second link, but that answers this pretty well (the second page, iirc). Player physics in games are pretty far divorced from reality -- and for good reason. They simply don't work very well for games. Consequently, you'll find no end to the articles discussing the design of player physics. Common to all of them, as mentioned earlier, is the need to iteratively adjust various values until things "feel" right.
Again, this is a situation where a strong maths background seems essential, but really isn't. Basic arithmetic, and a good aesthetic sense, is sufficient.