Regarding the controls, if you intend to provide multiple control mechanisms, you might consider having each control mechanism be exclusive from the others, and re-use the same set of basic keys for each, rather than proliferating keys: being faced with a long list of keys, even if further study reveals that they belong to separate mechanisms, can be somewhat overwhelming.
You also mention that you like this particular control mechanism; have you seen (and enjoyed) it in other games? If not, perhaps it feels so intuitive to you in part because it's something that you created: you've been using it since you started implementing it (presumably), have learned its use as it has developed, and understand it more intimately than a new player might.
However, again, it may well just be that it doesn't work for me: as I think that I said, I recall having similar issues with other prototypes that used similar mechanics.
(For myself, I think that I'd prefer that WASD move the character relative to the screen, and that the character attack either the indicated character (if one is clicked on), or in the direction of the mouse.)
Not sure about camera speed, maybe smooth them up?
The rotation seemed smooth enough to me, as I recall, just far, far too fast. Note, however, that this refers only to the rotation, not translation--the latter seemed fine, I believe.
For 'Auto-Attack' it will make you attack target when you face them and in melee range, which should allow players to focus on movement, but manual attack have advantage of anticipate enemy movement, like you can attack before move into range, or aim front of their moving path.
Fair enough; I don't think that I ever managed sure enough control over the mechanics--nor had enough reason against the enemies as they are now--to attempt such a thing.
Regarding the enemies, what about behaviour that's slightly more complex: have them attempt to flank the player, or otherwise avoid being kited.
