Drive isolation transformers were around long before the widespread use of reactors. It is an isolation transformer because it has two separate sets of windings on a common core, so there is no direct electrical connection from one side to the other; they are only magnetically coupled while being electrically isolated. Drive isolation transformers ARE also typically delta-wye so that there IS a phase shift from primary to secondary, even if the ratio is 1:1. This is because drives, both AC and DC, are typically designed for solidly grounded wye power systems. So you need that drive isolation transformer if your source is NOT a solidly grounded wye. Reactors on the other hand are a direct connection from line to load, there is no separation.
From the standpoint of mitigating harmonics, the transformer and reactor will have about the same effect, but reactors are typically available in different impedance values, where as drive isolation transformers are not, they are usually about 3-6% impedance, but not really selectable in that way, you usually get what you get when you buy them based on the motor size.
But electrically, an isolation transformer is basically equivalent to two back-to-back reactors, or look at the other way, a reactor is half of an isolation transformer. From a standpoint of their main benefit, that of slowing down the rise time of transients, they are basically the same. Adding either one ahead of a drive adds an inductive time constant that will not allow a rapid change in current or voltage through the coils. That's what you want. Shielded isolation transformers, as most drive isolation transformers are, also add the the ability to isolate common mode noise created by the drive from getting back onto the incoming line and affecting other equipment. Reactors can't do that.