Initially was under the "Experimental" category, I went through the code, cleaned code & fixed pgr_edgeDisjointPaths that is why there is a:
Breaking change on:
Added path_id, cost and agg_cost columns on the result
Parameter names changed
The many version results are the union of the one to one version
And I moved them up to the "Stable" section.
On version 2.x only the the functions on 2.0 are in the main area of the documentation, everything new either is "Stable" or "Experimental"
The Stable ones are the most likely to be moved to the main area of the documentation on version 3.0
Everything done by GSoC Students is on experimental section unless there has being a procedure like the one I did on the max flow functions.
And, like Daniel mentioned, feedback is important.
And you are not on your own, i've being feeling that way for this last 2 years of code rewrite, but now I am into a point where I think the internal code is stable and more members of the community can participate in bug fixing, more rewriting, or adding new functions with ease. For example, currently, I am on gitter , I am still working with Vidham, and Anthony Tasca, with the focus on the functions needed to make pgr_trsp work correctly.
A way to learn how pgRouting works internally & contribute would be to read code and document the code. (Just by reading 2.0 code I knew it had bugs)