I don't think this is about 'stance' or 'policy' as you suggest. Simply, MS from the get go - before release od X1 - had a vision of unified Windows platform with Xbox as a gaming brand, where we'd access the games via PC, Xbox Scorpio, X1(S) maybe even Surface tablets and phones; and the experience - resolution, framerate, level of detail, etc. - would be scaled up/down depending on your hardware. That could also explain why X1's development environment was such a mess - compared to that of Sony's, allegedly "a hardware company"! - at the beginning, simply because they were aiming for a much higher goal and there probably were delays with DX12, Windows10, etc. Because of that plan the devs were required to use higher-level APIs to make their games, e.g. if game wants to draw a line it'll say to the operating system "draw me a straight line from coordinates X1/Y1 to X2/Y2 in color RGB".
Sony, I think, came up with the idea of incremental updates later in the process and many games already were using their more low-level API (called GNM, I think?) to get more from the hardware, so in PS4 case the game instead of saying "draw me a line" would actually order the GPU to light up' individual pixels in the graphics memory buffer.
That (very, very simplified example!
) could make the PS4 games run faster (because there's no middle man interpreting the orders), but also made the software very much dependant on the underlying hardware. Xbox's "interpretation" layer allows to more easily swap the GPU or increase its clock and see the benefits right away - the software patch would only have to be done once, on the system / interpreter level. Whereas for PS4 it would probably break things, unless specifically patched for each game.
So, as I say it's not a matter of stance or policy - simply in this case I think MS was smarter and more forward-looking.