Jotaro, if you keep calling yourself a fan, I might actually start to believe it. :) But thank you anyway for the gesture.
I'm travelling, and the computer I brought along is embarrassingly slow; so I didn't play past the first few levels or far enough to yet find any cognitive bottleneck(s). I'll try to play it again later.
I can't say for sure what will work for you as an individual, so this might be a good opportunity to personally experiment. I would hate to waste your time. I can, however, offer a few suggestions which I believe might work.
Many VS learners and thinkers, myself included sometimes, fumble through everyday tasks in an entirely unmolested associative and slipshod way, having a hazy final objective in mind but taking few pains to methodically think through upcoming sequences. Basically, this person allows memory to conduct his or her almost every action around the clock. Overriding (and overwriting) these habits, of course, means doing something which is fundamentally different. Namely, we might do exactly the opposite; we might analyze as thoroughly and meticulously as possible, or as the situation calls, as thoroughly and meticulously as needed.
I stopped and considered the other day just how many steps there are involved in even the simplest of tasks when I had been thought streaming and decided to apply the same procedure to fueling a vehicle - which I happened to be doing at the time. Even with such an easy task, one that I would readily entrust to a child, there are surprisingly many steps if one cares to think about it in detail, enough to quickly overwhelm the working memory.
In the same spirit, I would try, as an exercise in building sequential processing skills, mapping out all steps of all actions I plan to take completely and ahead of time in thorough detail, then executing the task using mostly content of your short-term memory. This doesn't have to be as rigorous as the thought streaming process, though I earnestly believe that to be the best solution. Just a few steps ahead in much more detail than accustomed for any of the simplest everyday tasks will do, but I'd start slowly. I would be constantly working to extend the scope and depth of this process as I habituated. This effort requires analysis and forward thinking, critical skills that are sorely wanting in many persons with this style of thinking. It helps to break the spell of automatic thinking.
Also, IIRC, in one of his last few posts before the recently disturbed hiatus from regular posting, argumzio recommended in a thread on LSAT reasoning prep not to be afraid to posit a solution, to guess, and to be able to push forward when an idea doesn't work out; I'd echo that recommendation. It's especially pertinent given that the issue of perserveration is only due to a consistent and ungainful repetition in problem-solving efforts. Awareness of this inclination toward repetition would likely help a great deal. I would take each instance of perseveration as an opportunity to reinforce the decision to think consciously. Each time anyone makes an error of any sort, in my opinion, is a golden opportunity to learn something about his/her thinking processes - where s/he erred, what could have prevent this particular error, and so on. With enough vigilance, and practice consciously analyzing, the problem will probably sort out itself.
While I think interventions based on short-term storage and the predictable manipulation of random stimuli (e.g., n-back tasks) are useful for fixing symptoms characteristic of atypical processing styles and disorders with common working memory deficiency (ADHD, dyslexia, etc.) in their own right, this has, perhaps, more utility still because of the focus on decision-making procedure and the modest practice in logical deconstruction. And the assumed reduction in day-to-day careless errors would be nothing to dismiss either.
--Brandon