I find this unconvincing. First, it is true that marriages are based somewhat on income similarities, but I think this is due to social circles generally being so shaped, not on calculations that might drive people to search outside these circles. Since the tax change is not affecting the class makeup of social circles, I don't see any reason for less income mixing within marriage.
Second, even if your calculus of marital choice were correct, it doesn't seem to me that the income changes from the tax changes would be either very great or terribly salient to people considering marriage; I would like to see the salience of partner's income compared to other characteristics, and more importantly the salience of tax laws as they affect choice.
Bill