It is my opinion that experimentation should be transparent to the listener. By this I mean, I do not want to subject the listeners to everything I experimented with that did not work. This also implies that I have a certain aesthetic - I want to produce things that I would want to listen to -- and I do have a preference for at least some traditional melodies and rhythms. I don't mind, however, if it takes many listens to fully grasp what's happening.
It took me a few years of listening to various Miles Davis albums before it just clicked for me. Bitches Brew is a very challenging album, but once the familiarity is there, it's something you crave! And that can be a very satisfying musical experience. Same thing with King Crimson - at first I thought it was just arty noise, then once I acclimated to what was happening, it all made perfect sense.
IMHO, some fans of "experimental", or "avante-garde", or, ahem, "true" progressive music, are really just people who enjoy more dissonant melodies, or to whom repetition can become easily irritating. If we review the ingredients of what people consider to be experimental:
- Melodies that avoid predictable structure - in doing so, actually adhere to structures that are just as deliberate as the traditional ones by using sharp 4th and 5ths, flat 9ths, microtonic scales, etc.
- Rhythms that avoid traditional rhythm - inevitable follow predictably disjointed-sounding meters, polyrhythms.
- Avoidance of repetition.
So, really, the experiments were done by composers hundreds of years ago, anyone employing these devices since the first person to try it is no longer experimenting, they are simply rehashing dissonances used many times before - it's music that is only unpredictable if you expect it to be melodic, or have an easily discernible pattern.
Now, the real genius is with composers who can avoid conventional devices, but do it in such a way that no listener can tell...But nobody would ever categorize that music as experimental. Because the experiment worked!