When I look for a definition of eugenics there is wide range out there. Most of them seem to deal with selective breeding, which is what I found so shocking about his comment. I guess I would have to hear stePH's definition of eugenics. If it is "some people shouldn't be allowed to breed" then I have a serious issue with it for several reasons. If stePH's take on eugenics is, as you pose, reducing susceptibility to disease through genetic manupulation of fetuses that's a different matter.
First, with regard to eugenics, I find it a rather dubious notion. The "success" of breeding programs for plants and animals comes largly from a very narrow definition of success. A North American wild turkey is clever (for a bird), adaptable, quick, strong for its size, and tenacious. It's domesticated cousin can't fly, can't reproduce without human help, can get lost in a farmyard, and is generally acknowledged as the dumbest of the common North American domestic animals. (Barely beating out sheep, another intensively-bred species.) However, it's considered "successful" because it will generate more white breast meat per pound of grain in a shorter time than its predicessors.
I like Thanksgiving dinner as much as the next guy -- maybe more -- so score that as a "win." But I can't imagine applying a similarly narrow definition of "superior" to humans, and in the absence of such a narrow definition, I can't see eugenics doing much in the way of "improvement." One of the Iron Laws of Genetics is that improvements in a selected-for trait come at the expense of non-selected traits -- like our friend the turkey, who's good for producing breast meat, but not much else. And the more traits you try to select for, the longer it takes to reach a goal, and the harder it is to overcome the "friction" of regression to the mean.
When you consider a propostion as complex as what it takes to create a "good human" -- I believe that's been debated in the West at least back to the time of Plato, and probably much further -- I have trouble with the notion that eugenics offers much hope in that direction. Especially when you consider that the genetic component of important traits like empathy, intelligence, emotional flexibility, etc. is largely unknown.
For these reasons, I think "eugenics" in the sense of trying to control the traits in a population via conventional breeding is a dead end, and always was.
All that being said, I'm sure we'll deploy some of the newer, non-conventional genetic tools as soon as we think we're ready. Which will probably be way sooner than we actually are. Since there is probably going to be widespread acceptence of the idea we should "fix" or prevent disease and disability, that's probably where we'll start. Over time, as capability improves, the definitions of "disease" and "disability" will become increasingly elastic. I'm sure that something like my total inability to carry a tune will eventually be defined as a "disability" in need of correction -- at least if it has a genetic component and a fix is available. We're certainly seeing that today with things like height, with some parents seeking drugs originally intended for people with severe disorders to boost the height of their kids from say the upper 5-foot range to over six, because they think it will improve their prospects in life.