Runs fine at idle but hesitates occasionally

See if you can measure the three voltages on the MAP sensor with it plugged in and the engine running. You can usually back-probe through the rubber seal in the connector with a paper clip touching the terminals, one at a time.

One terminal should have 5.0 volts, one should have 0.2 volts, and the middle one will be between 0.5 and 4.5 volts depending on intake manifold vacuum.

Check the vacuum hoses to the MAP sensor and make sure they are hooked up correctly. Those plastic lines run back to the throttle body.

You need to be more specific than that. What wires were you reading from and what were the readings under what amount of vacuum?

That is not how you measure that.
With the key on, engine off you should measure between the black and the gray wire and you should see right around 5 volts. If you don't have that, then you need to test the wiring back to the PCM and find out why. If you do have 5v reference, then you measure between the black and green wires and slowly apply vacuum with a hand held vacuum pump and you should see a varying signal from just under 5v down to under 1v if it's working properly.

Start by suspecting a vacuum leak. That will cause a high idle without a corresponding increase in power. 17" of vacuum is lower than normal, hence the suggestion of a leak. Pinch off various rubber vacuum hoses to see if one causes the idle speed to drop. If it does, follow it or its branches to the leak. If you suspect a leaking gasket, use a spray bottle while the engine is still cool to spray water on them. If you see the water get sucked in, or the engine speed slows down, look closer in that area.

You can also get a clue with a scanner that displays live sensor data. Wrenchtech will set me straight if I'm wrong here, but since the idle speed motor is the same part that is used on Chrysler engines, (it just goes by a different name), I'm assuming you can check its operation the same way. On the Chryslers, the scanner will display the number of "steps" the motor is being commanded to by the Engine Computer, from step "0" to "256". For a properly running engine, step 32 is typical.

If you find it at a higher step, say "50", for example, the computer is raising engine speed in response to something. More commonly you're going to find it at step "0" meaning the computer knows idle speed is too high and is trying to bring it down but without success.

The idle speed motor could be stuck too. We often suggest that but it's more common for the air passage it controls to become plugged with carbon resulting in a low idle speed and / or stalling at stop signs. Here again, being more familiar with Chryslers, that motor can be run up and down with the scanner to check its operation without having to take anything apart. My guess is you can do that on GMs too.

I should have added, if that idle speed motor isn't able to turn freely to adjust the air valve, the computer will still THINK it's adjusting it. It will keep bringing the number of steps down all the way to "0" in hopes of getting the idle speed down. No diagnostic fault code related to the motor will be set in memory because electrically, there is nothing wrong. That's what the computer checks for.

Oops. I just checked the code descriptions for myself and got "35 - idle air control sensor circuit fault". I see where some people get confused with the operation of that motor because it is not a "sensor" which is a computer input. It's a computer output. The fact it's setting a code suggests an electrical problem.

That motor has four wires going to it and four electromagnetic coils between them inside. Pulsing those coils with varying voltage and polarity is what slowly turns, or "steps" the armature of the motor to various settings. If any one of those coils is open, (has a break in it), or if any one of the wires going to it is open, or if any one of the connector pins is stretched or corroded, the motor won't turn.

There really isn't an easy, fast way to check the wiring except by removing the plug from the computer and doing resistance checks. You can't test anything with a voltmeter either. Since there isn't much to check, the best approach is to physically check the pins in the connector, and while it's unplugged, measure for continuity between all four pins on the motor. I have no clue what the normal resistance should be. All I would care about is none of them are open. To add to the confusion, if one of the four coils is open, you'll still measure continuity between all of the pins by measuring through all of the other coils. I don't think the computer will even detect one open coil. For an open to be detected, it would probably have to be a break in the wire harness going to the motor, not in the motor itself. As a result, substituting a new or known good motor is a faster way of finding a bad one, but be careful with the two mounting bolts. They're usually torx heads that strip and round out very easily, and to help that happen there is thread lock sealant on the bolt threads. Often it's hard to get squarely on the bolts too. Be sure to use the proper size torx bit with plenty of force to keep it straight and engaged in the head.

Another way of inspecting the idle speed motor is to unbolt it, pull really hard on the valve to extend it, then reinstall it that way. That will block the idle air passage and cause a really low idle speed. If the circuit is working properly, that valve will retract when the ignition switch is turned on, (Chryslers), or when the engine is stopped, (GMs), in preparation for the next engine start-up. (I might have the GM version backward; it might be the same as the Chrysler operation. The difference is you can usually hear that motor running right after the engine is stopped on many GM trucks). Regardless of when that valve moves, if it's working properly, you will get a nice idle "flareup" to 1500 rpm for a few seconds right after starting the engine, then it will drop right back to the normal 800 or so rpm. If it does that, the motor is working, the computer has control of it, and the air passage is not blocked with carbon.