Well, in my particular field (research in Academia) I've been on both sides of the job interview. The sad truth is that the two sides have somewhat conflicting interests, and this can't be avoided.

The person looking for a job will usually have some personal plans -- grad school, a family, perhaps simply personal growth -- that may decrease what he can contribute to his work. And the employer is looking for the person who can give the best results, regardless of other qualities. So both sides will try to fool each other. I suppose it's humanly impossible to have it any other way. And note that neither employer nor prospective employees are 'evil': they are both operating within the constraints of the system in which both are embedded.

The use of statistics is indeed relevant. Note that insure companies use it to evaluate how much a client should pay. Risk groups can be statistically determined -- of course it's possible that a non-risk group client will end up needing incredibly costly treatment, and if too many of them do this the insurance companhy will go bankrupt. But statistically -- and the hard data are there for anyone to look at -- it's the people in the risk groups who are more likely to need more costly treatments, and so it is (statistically if not humanly) fair that they should pay more.

Of course a man in which a lot of time and training was invested could also leave to go to grad school or to another job; he could also die in a terrorist attack, or suddenly convert to Scientology, etc. etc. etc. But what do the hard statistic data say? Of course, we need the numbers here. I don't have them, so I'm assuming results that I haven't myself researched -- so if this is not the case, please correct me and supply the appropriate data. But if it is true that women are statistically more likely to give up their jobs because of family (i.e., if the 80+ hour working moms end up being statistical outliers, anecdotal exceptions), then, all other factors (competence, experience, etc.) being equal, it makes good sense to prefer hiring men. This is not discrimination per se, it's simply acknowledging the fact that, in the jungle of the marketplace, it pays to get the odds to favor you as much as possible.

Again I say: this all depends on the statistics. If women are really statistically more likely to end up frustrating their bosses' expectations because of pregnancy and childbirth... Again, if men are more likely to frustrate their bosses' expectations because of a higher probability of dying in car accidents, or of going to grad school, then they should also be avoided. All we have to do is get the statistical data on these two cases and compare them. In fact, consider all kinds of possible futures (depression, midlife crises, sudden desire to move abroad, etc.) and study them statistically. Add all that up, and then decide who you'd rather hire. That is what insurance companies do, and they have to do that to stay in business. If it ends up being true that men (all other factors -- competence, experience, etc. -- being equal) are statistically more reliable by a significant margin, an employer would be simply stupid to ignore that. And the same is true if women end up being statistically more reliable: employers should prefer them (and if this is not the case, then you have objective evidence for prejudice affecting hiring practice: employers acting against their best interests because of hidden assumptions that favor men over women).

Without such data, what are all the cases described here but simply anecdotal evidence -- like my 80-year-old uncle who always smoked and yet never developed lung cancer? Still it makes sense for insurance company to consider smoking as a minus when deciding how much a given client should pay.

I have heard about anti-feminism backlashes quite often, but frankly I think this is a bit overstated. Old battles do have to be fought again; this is not so only in feminism, but pretty much anywhere. Look at Darwinism. Look at any Academic discussion. You hear old arguments coming back with a new life, and the discussion starts again. Some people even develop a pessimistic opinion about the whole process and think that progress is impossible.

I don't think so. In well-meant discussions, arguments are always worthy of attention. Even when an 'old battle' comes back to life, the new defenders won't state the old arguments in the same way anymore; there's always something new coming up. It's not really a circular, cyclic motion in time, it's more like a spiral. So anti-evolution activists like e.g. Intelligent Design proponents do reopen old battles, but their challenges are differently construed and, even when still philosophically equivalent to the more primitive versions previously dealt with, do bring something new to the discussion (e.g. the case of the bacterial flagellum). They are still wrong, but it is worthwhile to show that again. Always thinking and rethinking the foundations of evolution theory because of new challenges -- or because of new versions of older challenges -- actually helps it become stronger. (And should some challenger someday actually present a conclusive argument against evolution, indicating that a new theory is necessary -- well, this is how science works. We should welcome evidence that there is some problem with currently accepted models of how the world works.)

So it is, I think, with feminism. New arguments, or new versions of old arguments come around; so what? Do the same points need to be re-argued? That's always good exercise, and if well done it will strengthen the underlying base. And should there be an argument that makes obvious that some previous assumptions of feminism were actually wrong -- then the argument in question is doubly important, because it will make some genuine improvement in feminism possible.

Just my 2c.