3 shifts is probably too cheap. If you are placing that maneuver on the troll, then the troll gets to save against it. Also, you need to beat the troll by at least 1 to make the Aspect "sticky" and available to be tagged in future exchanges.
In this case, it sounds like a maneuver which the troll might resist using Conviction. Probably not the troll's best skill, unless it is a spellcaster or has other reasons to be particular resolute. So a lucky spellcaster's roll against an unlucky target's roll might make a simple 3-shift effect work. But to guarantee success, the caster has to overcome the target's best possible resistance roll, plus 1 to make it sticky.
If the troll has Conviction at +1, then the Maneuver should probably be at least 6 shifts (in case it rolls a 4) and possibly more (if it stands a chance of invoking any Aspects to increase its roll).
But, yes, as far as how useful Aspects can be, there can definitely be a wide variance in potential effect, and I think that rewards creativity.