Universal Acid suggests that "if you're in favor of progressive taxation and national health care, consistency requires you to favor an emphasis on educating the least-gifted." But that's assuming the only foundation for political liberalism is philosophical egalitarianism* - which I should hope is not the case because that kind of egalitarianism is stupid and unjust: justice requires that we give equal concern to the interests of all, rather than holding the worse-off to be more important than everyone else.
* (Well, I guess he might mean prioritism instead, which evades the stupidity charge but still strikes me as unjust. Anyway, the important point is about the alternative bases that are being overlooked...)
A better basis for liberalism is provided by utilitarianism: due to the diminishing marginal utility of material goods, redistributing wealth to the needy could help maximize overall welfare. (If it didn't, then we shouldn't do it.) But this does not require blind commitment to equalizing outcomes in all arenas of life. It all depends on the contingent facts about what would actually be best for all involved.
Now, in the particular case of education, it seems to me that there are several considerations which count against neglecting gifted children.
For one thing, students with more potential could receive much greater benefit from the same educational attention and resources than would less able students. In this sense, less able students may constitute a utility drain, sucking up all our resources, and to little effect. Better to concentrate our teaching resources on those who can actually obtain a significant benefit from it! Just like our material resources should be distributed to those who can benefit most from them (i.e. the poor). This combination of views is straightforwardly coherent on utilitarian principles.
Secondly, intellectual excellence exemplifies an intrinsically valuable mode of human flourishing. To allow such potential to be squandered is, in my view, an incredible tragedy, which we should care greatly to avoid. (Yes, I have perfectionist leanings. So sue me.) By contrast, there's nothing particularly 'excellent' about being wealthy, or having expensive operations whilst others die needlessly from medical inattention. One can thus consistently combine intellectual perfectionism with material and medical egalitarianism.
Thirdly, anecdotal evidence suggests that gifted students need to be extended. The mental anguish arising from a failure to meet this need can exhibit itself in the form of bad behaviour, anti-authoritarian resentment, etc. Naturally, we don't want our best and brightest to grow up to be violent revolutionaries or the like. They're capable enough to cause a lot of trouble if that's what they really want. Let's not drive them to the dark side, eh?
(For some non-anecdotal evidence: "studies establish that up to 20 percent of high school dropouts are gifted." What are we doing to them?)
Finally, it might be thought that, perhaps after a certain point, intellectual talents exhibit increasing marginal utility, i.e. the benefit of set 'boost' is better for you the more you had to begin with. (This is distinct from my first point, that set resources might lead to bigger 'boosts' in intellect for gifted students.)
From personal experience, I think this is true of musical ability, for example. After a fairly steep raise in benefits from learning how to play some fun basic tunes, you get several years of progress which aren't really all that rewarding, until you finally get to play more and more wonderful pieces by Beethoven, Chopin, Rachmaninov, etc. I would benefit greatly from being just slightly better at the piano, whereas being slightly better at the flute (where I am much worse to begin with) wouldn't do much for me at all.
It's an interesting question what utility pattern intellectual boosts might follow. Plausibly, the steepest gains are to be made from picking up basic numeracy and literacy skills, so a focus on those that are struggling in these absolute (rather than comparative) terms might well be justified. But I suspect that after that, the difference between being moderately dull and moderately smart may not be so significant, whereas boosting someone from great to extreme intelligence might have a much greater personal significance.** If nothing else, this might be because we tend to become more emotionally invested in, and dependent upon, our particular talents. I play the piano more than the flute. Likewise, we might expect gifted students to end up using their intellect more than struggling ones. (They might choose to go into academia, whereas others prefer to become craftsmen or American President.)
** (There are difficulties in conceiving of a standardized quantity for the 'boost', though. Is there the same "amount" of intelligence between the two intervals I cited above? Is this even a coherent notion?)