I find that the decision of when and where to deploy mages, is a tricky one. Unless you are sending in a special purpose mage, to cast a single powerful buff like Fire Arrows, then a single mage is unlikely to have a profound effect on the outcome of the battle. So the tradeoff between uncovering more powerful magic (or staying in the arms race at all) vs making an army significantly more powerful, sometimes is very vague. Also, some nations have specific tactics that can make the use of mages a risky proposition unless you are strong enough to protect them adequately.
But like I said, it's tricky. Well used, a powerful cadre of battlemages can bring your armies to victory much more quickly, gaining you access to greater income, and more labs to rectify the temporary loss of research.
An interesting anecdote - the first time I tried to kill a player in MP, I was R'lyeh and he was Atlantis. My army crashed down onto his castle nearly unopposed (I had defeated his army 2 turns previous), and I sieged for a few turns, and then stormed. He had at least a dozen Kings of the Deep, and they flat out murdered my army. Regrouping, I sent the majority of my mages with my new army, with the belief that if I didn't kill him on the second try, the situation was moot (later it proved that the FIRST loss ruined my timeline, gotta love hindsight), and I arrived just as I completed research on Enslave Mind. Well, let's just say that I still had Kings of the Deep in my army at the end of the game.
The moral? Mages are only worth what they can accomplish. If you reach a new level of magical research, but lose the war..... well, those mages were worth nothing to you.
