This is a line of development that-- while clearly useful-- seems somewhat hacky and unpromising to me. While I agree that this is likely to yield useful benefits in the short run, it strikes me that fixing one's internal structure in order to produce reliably correct external actions without these sorts of hacks seems more promising in terms of long-term growth and skills.

About a year ago, I thought that lucid dreaming was a great path to rationality. While lucid dreaming is a great way to train the skill of noticing confusion, I no longer recommend it to people asking me for advice on rationality practice, because I think you hit the skill ceiling relatively fast and it doesn't particularly lend itself to further development.

I'm worried that this strategy falls prey to the same flaw-- while it's quite effective in the short run, I think that people using these methods will ultimately have to learn the internal solutions anyway if they wish to progress to more advanced domains. Therefore, it makes more sense to me to just start with the internal solutions.

(Of course, if you need rapid skill growth in the short term, this might well be a useful strategy to adopt-- just be aware of the downsides.)