That's a really good example because the ability to drive is probably the domain that most people understand and have achieved mastery.
I would respond that if the goal is to drive from 123 Main St to 456 Columbus Ave during daylight, dry conditions, in low to moderate traffic and at the posted speed limit, Danica Patrick would not have a discernible advantage over your mother - she may look better while doing it though. It would only be if dropped the flag and let them go as fast as possible that Danica's advantages would manifest themselves. Overlearning's benefits can be limited to certain narrow areas within a general specialty.
This also why I don't care much about the extreme outliers like Leatham or Danica. The advantages they are enjoying are not limited to hard work. The simple ugly truth is that world class, elite athletes are different from the rest of us.
There are also clearly degrees of benefit conferred by overlearning. A B-Class shooter and a Grand Master are both competent shooters and better than most but the Grand Master will do better assuming they stay within a USPSA realm. If we made both shoot High Power competition, the advantages the Grand Master previously enjoyed would largely disappear. While the Grand Master would have the advantages of overlearning pistol shooting, it would have little carry over to High Power shooting (or skeet)
I do think that a fairer example would be this. We take your mom and a 15 year old with a learner's permit. We then give them free reins to drive that course as fast as they want. The fact that your mom was comfortable with fundamental driving skills would allow her to complete the course at speed better than the 15 year old with the learner's permit.
Your comment about overlearning to a particular standard is interesting. Earlier I said that the standards we train to tend to define our maximum reasonable performance so you're correct in that sense. I would offer that someone who shot the same POST course on a recurring weekly basis, although slower, would retain a greater percentage of their skill when aroused than a new hot-shot shooter who was gifted with a lot of twitch muscle.
The only reason I'm even asking the question is that the literature is pretty damn certain that overlearning/automaticity insulates skill from the debilitating effect of high arousal states. If there wasn't a documented advantage, I wouldn't care.