There is another thread on this.  The general opinion is that carb control is too slow to meet the needs.  The right prop can do a lot about keeping engine speed steady.  You probably noticed that none of the IC engine guys thought they had a problem. 

As mentioned, this topic has been flogged, most recently in a thread started by yours truly.

The bottom line: there's lots of really smart, technically accomplished folks who fly PA, and no one uses 'artificial' help to keep the engine speed constant. This in spite of the fact that you can go buy a helicopter speed governor, lash it to a Will Hubin timer, add a battery and a servo and have a complete engine governing system just with off-the-shelf parts.  So there are basically three technical possibilities with governing an engine externally:

(a) it just inherently sucks.
(b) it can be made to work, but not as good as (or no better than) the current tricks that are done.
(c) it can be made to work, but the right combination of technology and technician has yet to make it work.

If you take a look at the thread "Active Engine Control", http://stunthanger.com/smf/index.php?topic=18991.0, you'll see a thorough review of (c), with a few remarks supporting (a) (case (b) is, presumably, assumed).

I put the word 'artificial' in quotes above because the PA world does use techniques to regulate engine speed already.  It's done purposely, one puts serious effort into it, it's quite 'artificial' in that one does a number of things to make it happen, and it's as real as real can get.  There's people out there who know far more than I do about this, but I'm going to give it my poor effort at a rundown, and hope that anyone who corrects me won't be too snide.

Basically, the idea is that you run your plane with an engine that's sized to work really well on the upward legs of the vertical maneuvers, then set up the rest of the system so that it's significantly detuned when it's on the level (and is perhaps detuned even more on the downward legs).  There are two ways that I know that this is done:

• Use an engine that's sensitive to mixture, run it rich, and use a tank that lets it lean out on the upward legs
• Use a tuned pipe that's 'too long' for level flight

When you hear about the Fox "4-2-4" run, that's an engine that's running so rich that it misfires every other revolution (called four stroke, although it's certainly not the same Otto cycle that your car runs), but that leans out enough on the upward sections of maneuvers that it goes into a clean two-stroke, then drops back to a four stroke in the level.  Other engines (don't ask me which ones -- I'm regurgitating received knowledge here) can be run at a really rich two cycle, but will gain enough power on the up lines to make for a nice consistent run.

Running a pipe is just as odd to the uninitiated as the "run it rich" business.  In this case you take advantage of a quirk in two-stroke engine operation and tuned pipes.  Tuned pipes use the acoustic property of a specific-geometry exhaust system to supercharge the engine.  The explanation is long and there are innumerable excellent web pages on it -- Google is your friend.  One extra-quirky thing about tuned pipes is that the length of the pipe has an effect on how sensitive the engine is to load. 

The obvious quirk is that if you make the pipe too short then the engine will be exquisitely sensitive to load, making it want to run extra slow for increased load.  This is problem for racers, because they want that short pipe so the engine will develop lots of power at high RPM -- but sometimes it's a struggle to get the engine "on the pipe".

Get the pipe just the 'right' length for the running speed and the engine will be fairly insensitive to load, making it tend to run the same speed for a variety of loads -- sounds good, right?

Now, put your "I'll believe anything" hat on: make the pipe "long" for the engine speed, and the more you load the engine the faster it'll go.  I remember reading about this effect in the magazines when I was in a long fallow period where most of my modeling activity was confined to reading and wishing.  At any rate, put such a pipe on the right engine, slap it into a stunter, and you'll get more engine power any time the plane slows down, and less when it goes faster -- which is just what you want.

So that's it -- are there "engine governors" used in PA?  Not to my knowledge, or to the people who responded to my queries.  But ask "is engine speed deliberately controlled?" and -- after you carefully define your terms* -- I think you'll find your answer to be "yes".

Personally, my plan for this is to

• Use the current methods for inducing IC engine power changes in competition until I
• Switch to electric power because I like the sound (or lack thereof) better and
• entertain myself with governor design, because it sounds like a fun way to waste my time

* Someone will disagree with me.  Then I'll say "do you run a 4-2-4 run, or a pipe?"  They'll say "yes", then disagree with me some more.