I thought the decision to withdraw was Hadrian's, and he wasn't chosen emperor until Trajan was on his deathbed...
Romans garrisons already left most of Mesopotamia when Trajan returned in Romania. I mostly agree with
@Agricola, with some differences.
The clientelisation of central and part of southern Mesopotamia doesn't seems to have been the immediate plan. Trajan didn't made Parthamaspates his client in the conquered territories south of Nisibia before the latter part of 116, after he had to deal with the reconquest of Northern Mesopotamia (incomplete reconquest, critically), meaning that at least for the time being, he was more confident with not over-stretching Romania, but was still opening the possibility of a wider provincialization.
Furthermore, while the unability to take Hatra was problematic, the loss of Douros-Europa really prevented Trajan to secure entierly northern part of Mesopotamia, which is another thing that IMO, made him withdraw his troops from central and southern Mesopotamia in the late 116, leaving his client more or less on its own.
..and the Romans succeeded to an extent in reestablishing control after the revolts started.
As they did already in 115 or in Armenia. Before another revolt took place, forcing Romans to yet again fight for what they had took the year before. At some point, it's not the capacity from Romans to conquer central Mesopotamia that I put in question, rather than their capacity to hold it against semi-guerilla warfare and the rebuild-up of Parthian army, especially when the empire have other issues (namely in Dacia and Brittania, not to mention jewish revolts).
As said above, I could see Quietus managing to keep Northern Mesopotamia in roughly the same extent Severians managed to, and it would bring interesting changes, but the military and political pressure on Romania was a real problem, and I'm not sure the eastern provinces could take much more of the cost of post-Trajanic expansion, at last for the time being.