To be fair, unlike the history of women's legal status, that isn't the sort of topic anyone in here is likely to have studied in college or read much about in the papers, unless perhaps for personal reasons they've been seeking out information on it for a while. And I'm sure, considering even just the few custody stats I quoted, that if I looked long enough I could find something (though probably not from the NYT Magazine) analogous to the Belkin article INDY quoted, suggesting that all the mothers-to-fathers ratio in custody awards "really" shows is that most men don't care that much about involvement in their childrens' lives, and are happy to leave the heavy-duty interpersonal relations and 'busy work' aspects of parenting to the ladies, themselves being 'naturally' better suited for the hard-driving competition and status-oriented world of the workplace (and besides, what kind of selfish guy would put his own parental ambitions ahead of his children's need for a mother's love)? Which, hopefully, is readily recognizable as just as unwarranted a set of assumptions as Belkin's.
This is, I think, related to the shortcomings of law as a social leveller often cited with racial issues. Again, I don't see it as a satisfactory answer to "Why not an ERA?", but gender socialization and therefore sex discrimination encompass a lot more than what affirmative action quotas and the like can realistically address on their own. Different learned social interaction and 'networking' styles, different received expectations about parenthood as an identity, alienation and resignation to something less than desired after repeated embittering encounters with people on The Other Side who don't appear to recognize these things even exist--those consequences (and various others) do add up, and they can adversely affect men as well as women, but it can be extremely difficult to legislate against them or quantify them enough to win a discrimination suit. Attitudes and expectations may change once the most obvious barriers come down, socialization differences may become less pronounced, but it's a slow and gradual process and often too easy to deem completed to any 'reasonable' person's satisfaction, especially if you're not in the historically most adversely affected group. Of course it also happens that people will manipulate a protected status and try to use it as a battering ram; but whether that's happening is often a highly subjective call, and that's where what the stats say can be helpful as a reality check, and a starting point for collectively analyzing why the initially expected outcome didn't occur--and reconfirming the desirability that it should.