Say I have two sales reps, each with his/her own private TiddlyWiki on node.js.
Neither sales rep has access of any kind to the other sales rep's TiddlyWiki.
However, each sales rep does have view access to one tiddler from the other sales rep's TiddlyWiki: other sales rep's list of customers.
When a sales rep saves a new customer tiddler, the save process updates (behind the scenes and upon any new/edit/delete of a customer) that sales rep's customer list tiddler (say "Sales Rep 1 Customer List"), a view-only json data tiddler, with each index being kind of like a comma-separated-value record for a customer.
The other sales rep (say Sales Rep 2) has a file in the tiddlers folder that is a symlink to the "Sales Rep 1 Customer List." So that sales rep has his/her own view-only "Sales Rep 2 Customer List" that gets updated (automatically and behind the scenes) upon new/edit/delete of customer tiddlers, and also has view access to "Sales Rep 2 Customer List".
So each sales rep has access to their own customer tiddlers, but also has access to his/her own customer list + the other sales rep's customer list, which can then be merged into a consolidated customer list.
Essentially, each private TiddlyWiki on node.js, hosts CSV-like json data tiddlers, and those data tiddlers are made shareable via symlinks, and all of the data tiddlers become components of a distributed database for querying/reporting/whatever.
Just throwing that out there in case anybody else geeks out on, as much as I do, this kind of architectural experimentation.