In the digital realm it is super easy to take your session and "save as" a new name. This gives you the option of retaining the settings/plugins etc and explore from scratch for comparison. A session config file is miniscule in size so isn't going to overload the space on your hard drive. Since you are learning, I recommend picking one thing (like compression) and really play with it one feature at a time to extremes to learn and truly understand what it does and how it affects a given track or bus. Once you have initially explored (in our example compression) move to another like eq or verb. This is how one learns efficiently. If you change many variables you have no way to know which change you made was the bad one and which was good and which was only the Grand Illusion. You're foolin' yourself if you don't believe it......... (/flashback)

Some folks spend hours or days on the snare; others are polishing their raw mixes even while tracking. Some start with kick and bass, and some with the lead vocal (if there is one). What Jack said is extremely valid - just don't forget you'll have your own workflow, and, to an extent, it is genre-specific. If your doing acoustic Chesky-stuff, perhaps you'll search for the ultimate guitar sound on its own; if you're doing rock-pop mixes, it's not only the guitar on its own, but how it sits in the mix (it can even sound pretty thin when soloed, for ex.).

Im still learning myself ( one year ), and i have found my process to be ( assuming you had a good tracking foundation with mic placement, performance etc. already done ). Listen to each track individually and compress to tame peaks and tighten, EQ each track only pulling out frequencies not relevant to track ( need to learn this as it clears mix for same frequencies battling each other, "muddy mix" ), than listen to whole mix adjusting levels of each track, once levels are somewhat desired add eq boost to each track where it needs for the "sound you desire" ( need to research magic frequencies so you know what instrument/ track benefits from frequency certain boosts ), than add effects meanwhile going back and forth between eq again to tweak how its sits in the mix with the effect (reverb is a effect) just added. Well now you have a basic mix and is ready to soft knee compress and EQ. Thats what i have learned for myself, and would love any advice/help that would benefit my whole mix by seeing what can be improved with process mentioned above.

While mixing, I'm learning to focus on one instrument and one effect at a time and doing comparisons like what Jack said. I'll set up two or more identical snare tracks in the same session and play around with the EQ for a long while(I'm still new) and compare the two muting each one seperately. I'll often go back and forth between what I think works best. Then I'll move onto the next effect and eventually the next drum always comparing. When I have done this with all the drums I may end up with up to 5 different tracks for each piece and pick the few that I think fit best together. Finally, I'll import what I think are the best drums into a new session and start the same process with the other instruments. Importing them into a new session still gives me the option to go back and look over my other options without running out of tracks.

STYX will never die

I seem to be doing the same, doing save as and going back and forth between processed and EQd tracks. However, it seems a little goes a very long way as i seem to go back to original tracks with very slight processing. Of course tracks with heavy peaks and transients like picolo snare need a fair amount of carefully dialed in compression, EQ, and verb to sit right in the mix.