Redo interface/handlers to accept the all-individual model

[kltm]

This is dependent on the completion of #76, the more critical backend part.

Redoing the interface to accept the all-individual model is listed separately as other clients do not need to wait for the low-level graph editor to incorporate all the necessary changes before moving forward.

Essentially, the first pass of this is to make the current interface grossly map onto the new model system, mostly by switching from the in-manager methods to the request_set request builder methods and using the manager.request_with() method.

At this point, the interface would be ugly, but consistent and usable by /expert/ users in our already expert pool.

The second level changes are to 1) "intelligently" fold the expanded individual sets into single visual units and 2) make editorial controls usable with this interface. The "exploded interface should still be available as well.

Once this is completed, the ticket is closed and the all-individual model system is ready for general (and expert) consumption.

[kltm]

Talking with @cmungall, we will mostly likely opt for a single release after the completion of part 2, rather than a rolling release schedule after the completion of part 1.

[kltm]

We now have the local dev example (gomodel:taxon_559292-5525a0fc0000001 vs gomodel:taxon_559292-5525a0fc0000001_all_indivdual) that renders as expected.

I suspect the best way to handle this is to still create and handle the interface with "edit" nodes and edges (edit.js), but be more clever about their creation, mapping, and event mapping.

Essentially, as a first pass without editing, we just need two paths through the edit node assembler.

[kltm]

This issue is now being addressed as getting the current editor system to work with the "exploded" all-individual models.

[kltm]

With this milestone, all previous editing is now possible using the old editor with the new API; just ugly and horribly. Onward to #114.