I've read several threads about G owners driving loaner vehicles that are noticeably quicker than their own vehicles that are the same model year.

And I've read explanations that some cars that are frequently driven aggressively, feel quicker and nimble because the ECU has learned to adjust to this style of driving.

Is this true? or Hogwash?

From what I've read on technical engineering websites, ECU's gather an abundance of information, perform a multitude of calculations, and provide an optimum output for the engine to perform efficiently for that given situation.
In other words, the ECU will make adjustments for that particular moment for optimum performance.

But does that necessarily mean that an ECU is learning a particular style of driving habit?

Let's layout a scenario:
Both cars are identical.
Car "A" is driven very aggressively for 10,000 miles.
Car "B" is driven less aggressively without WOT for 10,000 miles.
These 2 cars race one another in 1/4 or 1/2 mile.
Hypothetically, computers controlled the actual driving of the 2 vehicles so that each car was driven identically during this race.

Which car would win the race?

I'm pretty sure that in a WOT situation, the ECU is operating in a closed loop operation and the performance would be identical.

Anything less than WOT, the car is operating in open loop and I believe where the ECU learns to adapt shift points and part throttle fuel/timing changes.

^ +1

If this is the case, does that mean that mpg would be less?

Based on your driving habits, yes, the ECU could get more agressive, yielding less MPG.