I haven't read it yet, but at Amazon I've browsed Alon Ziv's book Breeding Between the Lines, whose thesis is that interracial individuals are more healthy and attractive, as evidenced by greater symmetry, which in turn is supposed to reflect their lower developmental instability. That is, interracials are presumed to have genomes that are better at dampening the effect of noise in their developmental system, so that the actual phenotype more closely matches the target phenotype.
But at the foundation of any chain of syllogisms must lie solid, or at least promising, empirical justification. Right now I'm working through Developmental Instability: Causes and Consequences, which covers the single phenomenon of Developmental Instability (DI) from a variety of perspectives -- quantitative genetics, evolutionary genetics, measurement of Fluctuating Asymmetry (FA), and so on. I'll be writing up a longer post on this for GNXP, but suffice it to say for now that the deceptively intuitive notion that "Lower FA = Lower Instability" turns out to be much muddier. This research program really took off in the early '90s, and by the time this 2003 volume was published, the initial enthusiasm became quite tempered. In brief, the research is pretty equivocal as regards whether or not DI reflects FA, particularly with respect to "good genes" models of sexual selection whereby choosers are purported to use a potential mate's FA as an honest signal of its DI. The latter includes the subset of studies that find an effect of symmetry on attractiveness in human faces. Later meta-analyses (including one published for the first time in this volume) show that symmetrical faces really are more attractive, but the effect size is not as large as was supposed.
An admittedly rough rule-of-thumb to follow when judging the merit of studies on the FA-DI relationship is that the earlier studies (before, say, 1998) tend to be less reliable, largely because of smaller sample sizes and lack of repeated, independent measurements of the subjects' FA -- the latter is especially important since the magnitude of variation in FA is well within measurement errors (i.e., most individuals are not pronouncedly asymmetrical). Lots more to say at a later point, but for now it's safe to say that, if interracials really are more attractive, it would only be due in small part to their greater symmetry -- assuming the latter hypothesis is true (there is currently no evidence for or against it).
PS -- the most convincing evidence that something extraordinary occurs with interracials is if their mean is above or below the means of both parental populations. However, this isn't necessary: the definition of antagonistic or synergistic epistasis (in other terms, "depression" and "vigor," respectively) only requires that the mixed mean be above or below (respectively) the value expected from additive genetic effects. So, if group A with mean IQ of 100 mated randomly and in equal proportions with group B whose mean is 70, then the mixed group's mean should be 85 if only additive (linear) genetic effects were at play. Suppose it turned out that the mixed group actually averaged 95 -- that would still be "hybrid vigor," since it's +0.67 SD above expectation (where IQ distributions have an SD = 15). Conversely, if they averaged 75, that would still be "hybrid depression," since it's lower than expectation. How this affects fitness is pretty difficult to guess -- many traits affect fitness, and in general the various races have their own strengths and weaknesses. For example, Northeast Asians score about 1.33 SD above African-Americans on intelligence and the myriad traits that are largely influenced by intelligence (such as SES), but they are probably 1 to 1.33 SD below African-Americans on height, and probably something similar for levels of gregariousness and self-confidence. And dancing ability, fuhgeddaboudit.
Fitness is always defined for a particular environment, so making general statements about how fitness is impacted by interracial status is unlikely to bear much fruit. However, testing hypotheses about in which situations interracial mating would increase, decrease, or have no bearing on fitness seems a safe bet for getting robust results.