Clarify data model vs unicode characters

[handrews]

This addresses #221.

JSON Schema does not care about unicode characters, only the
data model. Vocabularies SHOULD account for this by working
with data model strings instead of the original unicode
representation of numbers.

For example, define a value for the "format" validation keyword
that indicates that a string should be interpreted as a fixed-point
decimal if you wish to retain precision information in your data.

[seagreen]

Look 👌 to me!

If you wanted to you could add an example (perhaps about how this means JSON Schema doesn't distinguish between 1.0 and 1), but that's a judgement call.

[handrews]

@seagreen thanks! I think I'm going to leave that out for now, mainly because there are a couple of issues filed that may address how to work with such distinctions using format. So if people are interested I'd rather they took a look at the GitHub issues than saw something quite that definitive and left it alone.