Also the complexity of the fighter, only stems from figuring out which options (feats for the most part) are garbage. Once you have that sorted out actual play is quick. Your turn will never approach the decision tree and/or the time element of a wizard, druid or cleric.
Not just that. I must also find out which feats synergize well with one another. 3e appeared to reward overwhelming specialization in a single role, which means that it tends to be better to focus in a single area, rather than spread yourself out over multiple paths. So right from the very start, you should ideally have already mapped out what you want to do from 1st lv all the way to 20th lv, and stick to it rigidly. Because your class features (ie: feats) are front-loaded, screwing up your initial allotment of feats is tantamount to blowing up your entire fighter career.
That is the trickiest part to playing a fighter, IMO. You face the problem of failing as a fighter(or at least, the fighter you envisioned yourself to be) even before the campaign can even begin!
One such example is the "Jack be Quick" fighter build from the CO boards at gleemax, which basically involves a complicated (and some say, questionable) interaction amongst multiple feats to essentially give you 6 attacks for every 1 attack the enemy makes at you.
I agree somewhat with the assessment about spellcasters requiring a great deal of system mastery as well (in that you need an intimate knowledge of how best to use the various spells), but that is something you can explore during the actual game itself, so as pointed out, the downside to preparing wrong spells is not necessarily as bad as a fighter who has chosen the wrong feats. So the burden of "system mastery" on the wizard is arguably less.
Ahh...that sounds a little lame. May revise my post later after I have gotten some sleep.
