though I expect JT and OS to chip in
We can't help ourselves ..
Problem with Vmca/Vmcg is that it all happens within a fairly narrow speed band and, more perversely, the average airline pilot is never exposed to the handling differences either in endorsement or recurrent training. If you dabble with it, in the absence of training exposure, you run the risk of being bitten badly when the tiger's tail starts swinging much quicker than you can react to and hold on. Having said that, it's probably best to just stay away from these limits for the great majority of routine operations ..
Specifically,
(a) a rapid thrust ramp up may result in a short-lived thrust exceedance (overswing, if you like) - I recall one fatal investigation where this was indicated in the subsequent low level departure loss of control.
(b) roll/yaw couples may create undesired levels of excitement - which is why some of us like to expose upgrading crews to the situation if the particular sim is at all realistic in the vicinity of Vmc limits.
(c) if the sums are done correctly, the derated/flex takeoff should be fine without increasing thrust.
(d) if you do elect to increase thrust, do it steadily and cautiously, consistent with the threat circumstances. If nothing else, this minimises any dance floor action with the ensuing gyrations ...
if you push toga when you take off with derated since the vmca/vmcg will be higher than v2
That may not be the case on the day ... but, in general, the pilot has no real way of knowing just what limits are at play hence the need to tread cautiously.