Looking at the CNN exit polls, it seems quite clear that the Marshall/Newman amendment hit Allen coming and going.
Early polling indicated that African American voters were not significantly more likely to vote yes on the marriage amendment than white voters. And this was born out by CNN's exit polls which showed white voters at 58-42 and black voters at 56-44.
On the other hand, CNN exit polls also indicate that black voters made up 16% of the electorate in this election. If that holds true, it means black voters made up a higher percentage of the electorate than in any other statewide election except Wilder's gubernatorial victory(17%).
Since black voters are more reliably Democratic in partisan contests (of which the amendment was not one), the push by va4marriage to turn out more pro-amendment voters, including those from African American constituencies, certainly did backfire on Allen.
The effort caused more black voters to turn out than otherwise would have been the case, and when they did turn out to vote yes on the amendment, they voted for Webb.
That this was an important aspect in Webb's victory seems clear where, as here, turnout in many high Democratic performance precincts was, on the whole, lower than in high Republican performance precincts.
Put this together with the fact that The Commonwealth Coalition's GOTV campaign was focused on turning out anti-amendment voters in swing districts (35 to 65% Dem performance) and anti-amendment voters were, on the whole, two to three times more likely to vote for Webb than Allen regardless of party.
Thus, it is fair to say that the ballot question turned out voters both for and against the amendment who were less likely to vote for Allen than for Webb.
Ironic, isn't it?