I’ve achieved above average results in a few different task/learning domains in an audodidactic fashion. A couple of comments:
(1) If you are not just learning, but also figuring out how to learn, you have to resolve to “waste time”. The only other option is to find someone else who has also done the autodidactic part, and to have enough of a philosophical grasp of the learning domain to assess the degree to which their stated methods are responsible for their success, and then obey them for the long haul. The philosophical acumen is rare, and in modern culture the inclination to obey is almost zero, so most likely you are going to go the autodidact route, which again, means “wasting time”.
(2) The key is to waste time intelligently. So you have to have a rough idea of the appropriate time frame over which to test a method of work and a rough idea of a realistic goal at the end of that time period to use to assess your methodology. In my experience you need to be willing to burn at least six months being extremely disciplined with your method, before you can honestly evaluate it. Again, there’s a lot of art here, but any six months spent working systematically and consistently and then actively trying to evaluate for yourself how things went at the end of those six months is time well spent.
(3) You also need to figure out how to tweak your methods in intelligent increments. What I mean by that is that whatever method you use for the first six months of work should form the basis for your next six months. If you change methodology radically rather than systematically, not only are you disrupting whatever skill base you built with your first method, but you are robbing your mind of the structure that you need to actually figure out what works over time.
The worst case scenario is that you pay money for something, try it for two weeks, decide it doesn’t work, fall in love with some new ideology that completely contradicts your first method, pay money for that, try it for two weeks, and then rinse and repeat, with lots of idle periods and wishful thinking in between. That is wasting time.
The inevitable inefficiencies of your first year or two of work are not time wasted if you use them to develop a learning process that not only works for your current domain, but builds skills and character traits that you apply to the next task you set yourself.