A 9 range unit with a conversion always guaranteed means an army of 30 kts will lose to an army of 20 kts and 5 monks.
This is already the case. Knights fighting knights will give enough time for the monks to convert 5 of them, then 25 knights with heal support win over 25 knights without. In that situation the conversion time does hardly matter.
However, if you try to target the monks with your knights, it does matter. Either they're converted before they reach the monk, or the monks don't work at all and get slaughtered (as Carlos pointed out, the extremum conversion times are much more likely). Both results are unhappy solutions. They leave too little space for strategical decisions, both for the one with monks and his opponents.
If a monk starts the conversion process, I have no way of telling if I should run away quickly or charge at him - either might be wrong. On the othre hand, a player sometimes has to rely on using monks to defend his economy - for example as a pocket player against enemy conquistadores, not having access to camels. Monks are his only choice, but even when he uses them well, what the player gets is a complete gamble deciding over losing many villagers or converting the enemy's army in one stroke.
Of course there have to be randomly generated elements in the game. However only because it makes players adapt to a certain situation. A player can look at his map and make the decision that a drush is not feasible. With monks, the random number is thrown AFTER the player made his decision.