Docs should explain the zarr data model

[d-v-b]

The documentation should devote some space to a basic explanation of the zarr data model, (using zarr-python to explore instances of that model of course). In particular I think this page could be expanded on, e.g. starting with the very very basics about how zarr works (how chunks are stored, how metadata is stored, etc).

[MSanKeys963]

+1 for this.

I also think we can add some illustrations of the Zarr data model to the documentation. I have made some Excalidraw illustrations and used them in my presentation slides, and I'd be happy to add them to the above-mentioned page.