Just ran into one more detail describing the "general properties" of Apocynaceae which mentions:
Flowers: hermaphrodite
"Because the receptive stigmatic surface is located on the underside of the stigma, cross-pollination is the rule of the family. Self-pollination is absent."
Dioecy appears to be the exception, not the rule. As a negative, insect pollinators seem to be required. As a positive, its flowers appear to be quite attractive to insects, combining appearance, a strong scent, a pollen reward, and a nectar reward. Also as a positive - while it's not mentioned - one would expect that if it's evolved such a stigma design to avoid self-pollination then it's less likely to have needed to evolve a genetic process to prevent self pollination.
ED: Oooh, jackpot!
http://edepot.wur.nl/165045"self-incompatibility is common in Apocynaceae (Rowley, 1980") - Really? Well, there goes my theory...
(still reading)
Okay, here's what they report:
Wrightia tinctoria: requires an insect visit, but the flower will pollinate itself even if the insect carries no pollen
Catharanthus roseus: Autogamous, self-fertile without insects.
Plumeroideae:Usually autogamous
Tabernaemontanoideae are: Said to be always allogamous. Yet in the two species tested, pollen from a given flower appeared to germinate anywhere on the pistil and reach the ovaries.
Amsonia orientalis: Not self-fertile (no fruit set when flowers covered)
Vinca minor: Requires pollination, generally by honey bees, to get the pollen to a place it can germinate.
Pollination generally by insects with long mouthparts - bees, butterflies, some wasp species.
Unfortunately, they didn't check any of the more interesting species, and they didn't verify that fertilized ovaries actually develop fruit with pollen from the same flower. And when they write that self compatibility is rare, I don't think they mean genetic self-compatibility.
Hmm....