Here's a good one: idle hunt with a twist...

[sleuth255]

So, if I start the car after its been sitting overnight, then shut it down within a few minutes, on the next start the idle will hunt wildly, eventually killing the engine. Throttle has no effect until it hits 2K RPMs or so (ie I can't find a throttle position that can reliably hold any RPMs below that). This problem will continue until I apply "the fix". The fix is to get it out on the highway and drive it for about 10 minutes at highway speeds. Then, everything's good again. Subsequent cold starts are fine, and as long as I don't shut the motor off within a few minutes after a cold start there's no hint that there's anything wrong at all. When the motor is warm, I can start/shutdown to my heart's content without causing the problem unless its having the issue and I haven't applied "the fix".

No codes are ever stored. Disconnecting the MAF has no effect. Disconnecting the battery to cold start the DME has no effect. Even if I wait 24 hours, the problem remains. Only "the fix" corrects it.

I'm thinking this is mechanical. Something sticking until it warms up. IAC or throttle body are prime candidates in my mind but how to test without purchasing/installing a new part to see if it fixes this?

[WBAD530i]

Why not just tear down the entire intake side except the manifold and see what you find. My guess is you'll find some ripped boots and other various vacuum leaks. Check the lines that go to the back of the manifold as well.

Also take off your DISA and inspect it. Your flap could be completely broken.

[geargrinder]

Agree w/ both your takes. Sure looks like an ICV problem but that could also be related to leaks. If the ICV isn't doing the job of keeping idle smooth then the DME is using the TB to manage idle control also which is what causes the big swings. When the DME is in 'idle control' mode and bouncing like that, it won't respond to throttle inputs below a certain level. What might keep the ICV from being able to control idle might be failure but also could be a big old leak.

Why it won't happen if you don't shut it down for a second, well that I dunno, that is weird. Are your engine temp sensors reading accurately? If the DME thought the engine was actually warm already but in reality block was still cold, that'd cause an idle runability issue like that. Maybe an SAI issue? Thats also a thing that only runs on stone-cold starts.

[sleuth255]

Looks like my next step is to pull my ICV and have a look. Apparently they can gum up over time and need to be cleaned out with carb cleaner. That'll make it easier to check the rubber boots for cracks too.