For various reasons, I've long been amongst the doubters that genes "for" homosexuality will ever be found, and this is a position that isn't about to change, but what has changed is that for the first time I've run into a reasonable genetic hypothesis which would explain the universal existence of homosexuality across societies and periods - assuming said hypothesis were true, of course.
Several lines of evidence indicate the existence of genetic factors influencing male homosexuality and bisexuality. In spite of its relatively low frequency, the stable permanence in all human populations of this apparently detrimental trait constitutes a puzzling ‘Darwinian paradox’. Furthermore, several studies have pointed out relevant asymmetries in the distribution of both male homosexuality and of female fecundity in the parental lines of homosexual vs. heterosexual males. A number of hypotheses have attempted to give an evolutionary explanation for the long-standing persistence of this trait, and for its asymmetric distribution in family lines; however a satisfactory understanding of the population genetics of male homosexuality is lacking at present. We perform a systematic mathematical analysis of the propagation and equilibrium of the putative genetic factors for male homosexuality in the population, based on the selection equation for one or two diallelic loci and Bayesian statistics for pedigree investigation. We show that only the two-locus genetic model with at least one locus on the X chromosome, and in which gene expression is sexually antagonistic (increasing female fitness but decreasing male fitness), accounts for all known empirical data. Our results help clarify the basic evolutionary dynamics of male homosexuality, establishing this as a clearly ascertained sexually antagonistic human trait.
If you're too lazy, too pressed for time or simply too unfamiliar with the mathematics and population genetic terminology in the Plos One article to read it through, I suppose you could do worse than read William Saletan's take on it. Frankly, I mightily distrust Saletan's competence to opinionate on scientific matters in general, after seeing what a hash he made of sorting out the facts on the race-IQ "research" carried out by Rushton, Lynn and the other pseudoscientists who orbit around the Pioneer Fund. Still, even a broken clock can give the right time on occasion, and having read the paper through, I can verify that Saletan gets the gist of the actual study right, though I believe that some of the conclusions he draws do not in fact logically follow from what he's read.
While I think this paper provides plausible mathematical models to explain the widespread occurrence of homosexuality, a model is not the same thing as a well-tested theory, and the models examined in this paper are quite clearly too simple to explain reality: they assume that homosexuality is a trait under the control of either just 1 or 2 loci, each with only two alleles to consider, but if this indeed were a good description of the facts, one would have expected at least one such locus of large effect to have been discovered and independently verified by now, which is not at all the case. The paper's authors do a good job of providing tractable mathematical models which match their assumptions about homosexuality, but these assumptions are themselves of doubtful universality (e.g. that homosexuality occurs at low rates in all societies), and in any case there is nothing stopping us from drawing up purely cultural hypotheses which use similar reasoning to arrive at the same results; for example, perhaps there are peculiarities of family rearing style which foster homosexuality in certain families' males while leading their female members to want and have more children.
No theory which attributes the occurrence of homosexuality mainly to genes can explain why homosexual activity was so common in ancient Greece and Rome or amongst Japan's samurai class, nor can there be "gay genes" at work when otherwise straight males readily engage in opportunistic homosexuality in all-male boarding schools, aboard ships or in prisons; what is more, homosexuality is not the only variation on human sexuality which is found in any society of a reasonably large size one may care to look at, and with many of these other sexual preferences it is indeed reasonable to say that they occur at fairly low percentages everywhere, but no one is likely to rush to "explain" them as due to genes which are beneficial to fecundity in one gender and detrimental in another. With all of this in mind, to glibly pronounce, as this press release does, that this bit of mathematical modeling has established what is at most a plausible speculation based on doubtful assumptions, is to engage in irresponsibility of a high order.
Now, as for Saletan's conclusions, the first - "it implies natural limits to homosexuality" - is clearly untrue unless one begins with the same purely genetic premise as the Plos One paper, in which case all Saletan is doing is engaging in tautology. The existence of genetic causes for homosexuality neither precludes homosexual acts being teachable sexual tastes nor sets hard limits on the percentage of possible practitioners in any given society - if either claim were true, we would never have heard of its widespread occurrence in particular societies and castes (see, e.g. those I mentioned in the previous paragraph). That Saletan's conclusion is most unlikely to be true doesn't mean we ought to start worrying about Tinky Winky, of course, and the hidden presumption that the spread of homosexual activity is legitimate cause for worry is one which is itself open to challenge; still, this research paper simply doesn't establish that culture and environment are of no consequence where homosexuality (or any kind of sexuality, for that matter) is concerned.
Yet another of Saletan's conclusions fails to follow from anything in this paper, and indeed another trait he mentions in the following paragraph flies directly in the face of this particular claim. When Saletan (drawing on the misleading press release) says that "this larger phenomenon can't be dismissed as a disorder", all he is doing is repeating the p.r. propaganda without bothering to think for himself. In his second-to-last paragaraph, he mentions the study in the context of sickle-cell anemia, but while being a carrier of one copy of the gene responsible for that disease is beneficial in highly malarial environments, no one would dare argue that those having two copies aren't suffering from a disorder. There are solid grounds to say homosexuality isn't a disorder - e.g., that gay men are no more dysfunctional in any aspects of their lives than straight men are (at least when given the freedom to express their preferences) - but the suggestion that they are gay because of a gene making their female relatives more fertile isn't logically one such ground.
Thirdly, Saletan's claim that "the benefits aren't really confined to women. They protect society as a whole" is simply absurd: genes don't benefit "society as a whole", they influence the reproductive success of individuals, and while it may be legitimate to conclude that under certain mathematical models the existence of sexually antagonistic "gay" genes may act to buffer population decline when fertility falls below replacement rate, it does not follow that what is good for the entire population is necessarily beneficial to all or even most of its members, unless the fecundity of "society" in general is what its member generally glory in - and I doubt, say, Hong Kong residents or early morning commuters in London or Tokyo would number amongst such groups ...
In closing, I'd say that we are not a single step closer to establishing what effect genes may or may not have in determining male homosexuality, let alone in a position to draw inferences of the sort in the paper's press release or Saletan's summary. All we can say for certain is that we now see that under certain assumptions, the long-term existence of "gay genes" in a population is not a mathematical impossibility. That won't stop various groups from seizing on Saletan's popularization of this paper to advance political claims, of course, not least of which will likely be that everyone will now be forced to see the necessity of granting gays equal rights - as if "I was born this way" ever helped a black person in the Jim Crow south or a Jew in 1930s Germany ...
PS: A Slate forum participant has interesting criticisms to make of this research. Another that I'll add here is that it isn't even clear to me whether or not increased female fecundity would really balance out the reduction in that of males within a family, as men have a much higher reproductive capacity than even the most fecund women do, e.g., from the Guinness Book of Records,
Moulay Ismail, the last Sharifian Emperor of Morocco, reputed to have fathered a total of 525 sons and 342 daughters by 1703, and achieved a 700th son by 1721.
Given just how much of an advantage males can have in the reproduction department, there is just no good reason to believe that any putative "androphilia" genes wouldn't rapidly have been selected to display zero expression in male carriers.