Agreed.
Other negative aspects I see in IMRT are ultra-fractionization and increased importance of registration.
With such high dose gradients shifts that would be considered allowable in normal therapy now pose a threat to the very structures that IMRT is designed to protect... I saw IMRT on a foot, and every time the patient moved his arms the leg would shift by 2 or 3 mm - this made a significant difference in what was in and out of the treatment fields. No point in using IMRT at all, IMHO.
Also, as an engineer, with start-up ramp effects IMRT may lead to underdosing or overdosing. Instead of 25 treatments*6 beams = 120 something cycles, you now have 25 treatments*50 fields/beams = 1000 cycles, any systematic errors in the electronics will rear their heads. For this reason I have seen many places place limits on the number of IMRT fields, or a minimum on the individual field energy.
Perhaps its just immaturity in the technique, and all of this will improve with time.
Oh, and if peds is your area - have you been to St. Jude, in our dept?