Updated evaluation function docs

[m-messer]

Addresses #36 and provides other needed updates to the evaluation function user docs, to align with Shimmy details.

[peterbjohnson]

I've made a couple of comments. The first one is just in one place - I avoided repeating it - but overall in some places I was a bit confused by what's generic and what's python-specific.

In many places you have made it clear enough, but in a few it's not clear to me. Could you do a sweep from that perspective and see if your can make any improvements?

Separate comment: the overview of the python function is useful. We need to ensure no duplication, so the READMEs in the functions should point to this page, or vice versa. Will you/have you check(ed) that everything is consistent and with no duplication?

Finally, should we introduce a convention for what to do with the main README when a new eval function is created? Currently they all look generic - they should be changed to point at their own docs. Can we systematise that somehow? At the very least say here in these docs that it should be done?

[m-messer]

Thanks for the comments, they should be addressed in the recent commit.

[peterbjohnson]

Thanks all good. One typo I spotted: 'byt'. Otherwise, approved (can't see the button).