Incorrect. The easiest place to see where this breaks down is in backwards compatibility. Web browsers need to be backwards compatible with the vast majority o

Incorrect. The easiest place to see where this breaks down is in backwards compatibility. Web browsers need to be backwards compatible with the vast majority of web pages out there. That's half the point of the Web. If you define standards by implementation, you can never fix bugs once web sites start using them. (If you think you can just tell web sites to suck it up, consider the vast number of configuration web pages burned into ROM. And yes, they do use new features.) The standards themselves have this problem already, even though they're at a much higher level. (See Appendix B in the JS spec, for example.) The problem is far, far worse at the lower level of implementation. Relying on a single implementation will quickly produce the next IE6. And if you think Google programmers (or any company X's programmers) are good enough to avoid that sort of thing, then consider how well resourced those programmers will become once there's no longer competition or even a basis for comparison. Monoculture is bad, m'kay? (More detailed argument: https://blog.mozilla.org/sfink/2013/02/14/browser-wars-the-game/ https://blog.mozilla.org/sfink/2013/02/14/browser-wars-the-g... )