I agree 100% with grey-area. But the trouble with vector artwork is the lack of an efficient widely-accepted storage format. SVG is full-featured but not particularly compact, and it's difficult to write a rendering engine which handles all its features correctly. There is such a big impedance mismatch between what SVG specifies and what GPUs can render without a lot of CPU-side preprocessing.
Since font technology is so old and so efficiently implemented, the obvious thing to do is to extend the same technology to rendering color fonts AND color vector artwork, such as icons. But Apple is going the wrong way on that too: they added emoji support to their fonts by embedding PNG images. Yuck. This article has links which explain more about the MS proposal, which is much better: https://en.wikipedia.org/wiki/OpenType#Color The glyphs can be built with layers, each of which has a color and a vector path. This is what we need. And then we need the Khoros Group, and then the GPU vendors, to focus on accelerating exactly the kinds of vector paths which are found in fonts. This will give us a better solution for rendering fonts and also a way of rendering resolution-independent icons and other types of application artwork at the same time. And the font format will constrain the icon designers to create icons that are efficient for the GPU to render: as few layers as possible, and without all the SVG features that make rendering so difficult.
In the meantime, it's a nice coincidence that the vogue of the moment is going towards monochrome icons. That means that the use of icon fonts (such as Font Awesome, and others) can go on for a few years even without adding color. But we can depend on there being a backlash eventually: we have color screens and are using them to show monochrome UIs? So the technology needs to be ready by the time the designers get back around to that way of thinking again.