It appears that Mastronarde is thinking along the same lines as I was when I wrote "Maybe διδῷ is from διδώ+ηι, but Smyth 749 says that this form is from the weak stem, διδό+ηι." But to me, it seems like δηλοῖ is the anomalous form that calls for explanation, not διδῷ:
* The alpha and epsilon contract verbs have long diphthongs (graphically represented by long vowels with iota subscripts) in the corresponding forms of the subjunctive.
* The other athematic (-μι) verbs also have long diphthongs in the present subjunctive.
* Contractions where the second element is a long diphthong otherwise result in a long diphthong. See Smyth 59. http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0007%3Asmythp%3D59
* The contraction of ο+η yields ω, as in the 2nd person plural of the subjunctive of δηλόω: δηλῶτε.
But that's about as far as I'm willing to carry this exercise. This is one unresolved question that will follow me to my grave.
It has been useful, though, in forcing me to review the subjunctive forms of -όω verbs and alerting me to the fact that forms which I would "process" as 2nd and 3rd person indicative could in fact be subjunctive, when I encounter them.