I don't believe you
I'm not complaining , I'm all for Druids and mages getting a boost but there does seem to be some imbalance .
The perceived imbalance may be due to the way in which the term 'random' is viewed.
For example, consider I have a coin which I flip 10 times, and I record whether it lands heads or tails.
This is one possible outcome of such a trial (where H=heads, T=tails):
HTTHHTTTH
Now, imagine that I did another 10 flips, and this was the outcome:
HHHHHHHHHH
I imagine that most people would assume the first trial to represent a 'random' series of coin flips, whilst the second represented a 'non-random' series. This view of randomness is known as the Representative Heurisitc, and the basic idea is that people will view a certain situation based on a pattern of their previous experiences or beliefs about that situation. In this case, the belief that in a series of coin flips there should be no obvious order of heads and tails.
However, the probability of both series occurring is the same, as a coin flip is an independent event (which means that the outcome of a coin flip does not depend on the outcome of a previous coin flip).
There is a similar situation with Aggragoth, and it (along with the coin flips) is known as a Gambler's Fallacy. This is the view that the probability of an event (in this case, what drop will a player get from Aggragoth) will change based on previous events (what Aggragoth has dropped before).
If Aggragoth drops 3 Mage items, then a person who is relying on the Gambler's Fallacy will believe that the odds of Aggragoth dropping another Mage item when he is defeated will be very low. However, the probability of Aggragoth dropping a Mage item has not changed, and therefore it is entirely possible that the next time Aggragoth is defeated he will drop a Mage item.
I hope that goes some way to explaining why some people may see Aggragoth's drops as not being random, and I hope I've explained it in such a way that makes it (relatively) easy to understand.