Tony, as you may recall, I've experimented a bit with Joshua Fontany's great JSONMangler plugin and (nested) JSON data tiddlers. I think, the plugin provides the tool you need.
But, after a bit of a learning curve, I found I was able to get it to work and to access and retrieve data. I am not fully happy, though.
The main issue I have is that JSON only offers minimum features. It stores data. Period. When you want to retrieve data, they come out in the way you have stored them, unless you want to mimic features in TW normally offered by a *DBMS. E.g. for retrieval in a given sort order, you will need to either use one of the sort functions on the whole data set or create an ISAM feature in TW. As I wrote before, this is a rabbit hole I would be reluctant to jump into. There has been work put into this topic by minds more proficient than me, So I'd rather resort to the work done by them. Also, I don't know, how JSON performs with large datasets, but I would assume at 800 records there won't be too much performance issues.
So, for the time being, I will stick to my TW-based solution, but in the long run I feel tempted to try out PouchDB/CouchDB which seems like the natural complement for this kind of problem.
Best, Werner