To me the most important step of learning was using several commercial songs as a reference. Comparing the frequency curves with Voxengo SPAN I learned how the project must sound in my modest monitor speakers and headphones to be somewhat in the same ballpark as the commercial ones that sound OK in all systems.
What I still find very difficult, though, is getting the low end be reasonably audible (=harmonics) also in poor computer loudspeakers and such.
The differences between reproduction gear can be unbelievably big for a beginner. My latest lesson of that was a song in which I tried to produce a Tom Petty-style shimmering rhythm guitar background with acoustic guitars. It sounded quite ok through my Behringer monitors, almost good through my DT-770 headphones, but totally, horribly untolerable through my Sony Hi-fi headphones. Through the Sonys it was like scratching a zinc plate with a fork + transposed an octave. It was hard to believe I was listening to the same track.
I just had to cut down a lot of the (assumed) high freq air.