Add a way to limit the number of records retrieved by Kinto.js in one sync run

We can probably use the `_limit` param https://github.com/mozilla-services/syncto/blob/master/syncto/views/collection.py#L37

Read the referenced syncto/views/collection.py line 37 to understand the _limit parameter, then trace how Kinto.js builds its sync request. The work is done when a sync run can limit the number of records retrieved and the behavior is covered by the relevant tests.

Written by the indexing model from the issue text.