Nobody has tried to implement, only somewhat overlaps existing implementations, and probably contains some nontrivial set of computationally expensive or infeasible standards. (See quincunx's comment elsewhere in this discussion, for example.) Then people online complain how long it takes the browser makers to implement full compliance to, say, the latest CSS spec.
In general, the W3C has worked better than just letting random groups run rampant, but I am coming around to preferring the style where we don't call it a standard until there are at least two implementations already.
Microsoft: Here's a working, fully implemented and debugged, format for doing vector graphics in a browser. How about we standardize on that?
W3C: Cool, we'll take a look at it... [overthink overthink overthink] ... OK, here's our brand new, "Standardized" way of doing vector graphics in a browser, which nobody has tried to implement and only somewhat overlaps with IE's working reference implementation.
Microsoft-hating-developers (which was understandably most of them back then): Grr!!! Microsoft, you've broken the STANDARDS again!!! M$ Sucks!
I remember this exact cycle repeating itself a half dozen times in the 90s, and it took a lot of the fun out of trying to build web apps that worked cross browser.