The inclusion of με has stronger weight in this verse. First, it is the earliest reading (P 66 , from ca. AD 200) and has strong early and diverse support (e.g. B W D Θ Q f28 f33). That is to say, με is found across the Alexandrian, Western, Caesarean, and Byzantine text types, with a large geographical distribution. Its omission is thus later, more sporadic, and typical mainly of only the Byzantine text. It also is the more difficult reading. Its inclusion raises reading (such as the redundancy of "me ... in my name") and exegetical difficulties (such as its tension with John 16:23), which is an argument in favor of its originality than against it, as its omission eases these issues. It is still a subjective judgement call, of course, but the weight of evidence favors inclusion. What is interesting is that the NWT here departs from the W-H critical text (as seen in the KIT), which it does not on textual grounds but on exegetical ones. The footnote in the 1984 NWT shows that the inferior reading is preferred because it is "in agreement with 15:16 and 16:23". Contextual considerations like these may offer secondary support to a superior reading but are usually insufficient to overturn it (as the principle of lectio difficilior usually favors the more difficult reading), as writers are not necessarily consistent and ancient writings often are heterogenous in their redactional history. Indeed in biblical criticism ch. 15-17 are often recognized as problematic because they interrupt the connection between 14:31 and 18:1, so these may have been interpolated or dislocated in part from another location (e.g. ch. 15-16 displaced from their original location between v. 35 and 36 of ch. 13, with ch. 17 as a secondary addition). In any case, the desire for harmonization is what seemingly leads the NWT to reject the W-H judgment in this passage (whether or not there is also a theological Tendenz wrt Jesus as an object of prayer).