I received a lot of questions on this topic so I gave it some additional thought.  I think the question can be narrowed to the following: When (if ever) did the beneficiary become the "income tax owner" of the funds in the 529 account?

It seems there can only be two answers: (1) at the time of contribution or (2) at the time of distribution.

Is the beneficiary the tax owner of the funds at the time of contribution? While section 529 treats a contribution as a completed gift, section 529(c)(2) applies for "gift tax purposes." If Congress intended for the beneficiary to be treated as the income tax owner at contribution, it would have been easy to modify section 529(c)(2) to say as much. Moreover, general principals of income tax law would not treat someone as the income tax owner of property where the person (1) did not have legal title to the property; and (2) had no benefits and burdens of ownership (i.e., no risk of loss and, in fact, no economic upside or downside in the property whatsoever and no control over the property). The "income tax ownership" factors overwhelmingly support the conclusion that the account owner is also the tax owner at the time of contribution.

Is the beneficiary the tax owner at the time of distribution? Maybe? Probably? Does it really matter?  I'd argue it really doesn't matter so long as the beneficiary wasn't the income tax owner of the funds at the beginning of the tax year. If the account owner was the income tax owner at the beginning of the tax year and the beneficiary becomes the tax owner (because for example the funds are distributed to the beneficiary), then it would still be support from the account owner. Thus, I'd argue it's irrelevant who is designated on the 1099-Q and pays the tax (if any) on the distribution.

In fact, for those people who argue that the beneficiary is the income tax owner at the time of contribution, why would distributing the money to the account owner make a difference in the support test? For example, account owner contributes $20k to 529 plan. For argument's sake (and contrary to what I believe), let's say beneficiary is the income tax owner of the funds at the time of contribution. The funds grow to $30k at which point the funds are distributed to the account owner and the account owner uses the funds to pay the beneficiaries tuition. If beneficiary was the income tax owner of the funds at the beginning of the year in which the distribution occurs, isn't it still self support if the beneficiary merely gives the beneficiary's funds to the account owner who then makes a payment of the beneficiary's tuition? The intervening step of giving the money to the account owner to pay tuition has no substance whatsoever. We wouldn't argue for self-self where the parent takes $30k from parent's bank account and transfers it to student's bank account to pay student's tuition so we shouldn't argue for parent support when the student gives the student's $30k to the parent to pay the student's tuition. In my mind, who is distributed the money is a total red herring.  

Wishing the IRS would weigh in.