Had a chance to test opening the throttle while shutting down and it worked. It fired right back up after 15 minutes of resting.cornerslider is on it: The design and layout of the intake ports and the way the injectors are aimed leads to a lot of "wall-wetting" in the intake ports. Shut the engine down, then that all evaporates and fills the intake runners with fuel vapor. A restart right away normally isn't a problem, because it hasn't had time to all evaporate. A restart after a few minutes (heat-soaking the intake system) leads to the intake charge being too rich to fire for a few seconds of cranking (meanwhile the injector is dutifully adding more fuel as you crank). If it's been sitting long enough to be cold, it's okay, due to a combination of the vapors dissipating and mixing with air in the airbox, plus the engine being more okay with a rich mixture upon an actual cold start, plus possibly the idle-air-control-valve opening slightly in preparation for kicking up the cold idle speed.
Cracking open the throttle a little upon warm start encourages more air to get into the cylinder, to dissipate the excess fuel.
The bit about whacking the throttle wide open when shutting down the engine in anticipation of a subsequent warm start, is to encourage the last few rotations of the engine to happen without fuel being injected, to hopefully get some of that fuel hanging around in the intake ports, through the engine and out.
The wall-wetting is a pain to tune around, too. A certain setting of a PowerCommander, for example, at a certain RPM and throttle position leads to different air/fuel ratio depending on what the engine was doing leading up to that condition. It's not bad at full throttle, where most of the fuel that the injector squirts in goes into the cylinder right away, but it's a nuisance at idle and near-shut-throttle. The stock ECU has wall-wetting compensation (they all do, nowadays - emissions) but it seems like there is only so much it can do.