What I have found fascinating about ACS is the _lack_ of conceptual integrity, _and_ innovation, yet it soldiers along. Here's why:
Taking a look at the structure of ACS (particularly the 3.x series) it's fairly obvious that it was a dumping ground of ideas, mostly contributed back to the project after "lessons learned" during client implementations.
Not that there is anything inherently _wrong_ with that; after all, these modules were for the most part built for a specific application, and thus needed to prove themselves in the real world. The same can't be said for ACS 4.x, which has had a different set of issues since being released last fall.)
I'd also argue that the true innovation in the ACS came very early on in its development, ending perhaps in 1997, and was due to a handful of individuals like Ben & Philip. This innovation was limited to:
* the use of a relational DB back-end
* the common data model.
* making absolutely certain that the project used the most stable, proven platform available as its foundation.
Without these 3 conditions, the ACS - in any form - would surely have died by now. Everything since has been mere incremental improvement.
In fact, you can go to cgi-resources.com and download point solutions from other vendors for free (or a nominal fee) that best, feature for feature, every module in (open)ACS.
But you _won't_ get the common data model, and the preintegration that it implies. Or the reassurance that knowing no matter how crappily coded an individual module is [and acs has had some doozies], that the underlying system is fundamentally reliable and can be used in very demanding situations. Both traits that date back to the origins of ACS.
It is only relatively recently, with projects like phpgroupware, with its common data model, and the multi-threaded apache 2.x that people are people coming around to recognizing the benefits that ACS has enjoyed for so long.
Philip himself has called the problems that the ACS solves "uninteresting" at this point, and that they have been "solved". One can only wonder what the next "interesting" innovation will be....
