update: http://lesswrong.com/r/discussion/lw/mmu/how_to_learn_a_new_area_x_that_you_have_no_idea/

An excellent question! I have now written a meta document of "How do I learn X" which I have sent off to a friend to be checked before I post it on Lesswrong. I will post a link (probably in the next 24 hours) to a discussion thread. I think the meta-strategy is going to be more helpful and for more people. I can write a specific list after.

As a teaser:

1. Make a list of your knowledge in the area
2. Confirm what you think you know (via brief research), replace your existing knowledge with correct knowledge.
3. Make a list of the topics of knowledge in the area.
4. spend time finding the best resource (as a trade off of searching-time, $cost, time-to-cover-the-resource)
5. delve in, absorb the knowledge
6. devise ways to test yourself, and/or experiment that you have gained the knowledge. (experimental methods include controlled environments, repeated tests, evaluation)
7. teach others in order to cement your knowledge.

This list is a teaser; there are more points, but you can probably cover 1-3 by the time I get to publish the list.