Even if you "relegate" the system to vertical avoidance only (as above, whilst TCAS is pretty good at bearing info, it is fallible, and can easily make a safe situation catastrophic if believed), without an RA "decision" process - it can again lead to problems... TCAS uses "crossing RAs", where an aircraft above you, but descending, might ask you to climb... so a GA system where a pilot seeing an aircraft "above" (but not appreciate / judge it's high RoD) might descend and now create a collision.
I understand the point in controlled airspace, but in class G following your TCAS procedures (which are being related as a rote set of prescribed actions) is even worse. You have no idea what the other aircraft is going to do, as without the other half of the system, the other pilot is just going to use his judgement on where he thinks you are and make some manoeuvre (or just hold his course relying on the big sky)
ACAS is nothing like a controller since they are trying to provide different things. A controller is trying to meet separation standards, which are large.
In real life controllers seem to worry a lot about some separation but none about others. On a busy hazy Saturday in the South there are lots and lots of light GA aircraft, sandwiched into a thin slice of air being asked in some cases to navigate with positional accuracy similar to PRNAV (i.e. the gap between Biggin and Gatwick airspace - which at its tightest is only 3/4 mile wide on each side of the M25) with the warm words of comfort - 'limited information service only due to high traffic density and poor radar performance'. I am not complaining, only pointing out that the GA traffic advisory concept needs to be evaluated in a world of separation expectations much reduced to those of IFR in controlled airspace - and where high quality radar information is only provided on an adhoc basis.
There are numerous posts where some GA pilot says 'I was on a RIS and someone crossed me 50 feet above me', controllers respond with 'well it is only information and you are supposed to be looking out!' (which of course is true). Having a bleeping things saying 'traffic closer and closer same level' is a huge step up from 'Sorry, was busy'.
Most of the arguments I have heard against GA traffic detection systems have been based around a logic that most things out there to hit aren't visible to these systems anyhow and having one may give a false sense of security. The fact that it is not as good as having TCAS, turbines, and Class A from the ground up is true but is not a good reason not to take the first step. (It reminds me a bit of the logic that GPS approaches are not perfect so we should stay with NDBs
)