`<featurecaption>`'s Content model, Content attributes, and Contexts to be used

[Malvoz]

I believe <featurecaption>'s Content model should be the same as HTML <title> (Text) and not <figcaption> (Flow content). This is mostly to align with the title element because I think we'll want featurecaptions to behave similar to title elements (see polyfill issue on native tooltips) so that we can have tooltips on hover (and focus, as some browsers are working on) which is important such that featurecaptions are accessible to visual users as well as non-visual users using ATs.

The Contexts in which this element can be used should be "As a child of a feature element containing no other featurecaption elements".

Content attributes may as well be Global attributes (same as for <title>).