On 2015 Apr 26, Donald Forsdyke commented:

The data in this interesting paper seem not incompatible with the hypothesis that the need to prevent recombination with other organisms drives a organisms GC%. This anti-recombination selective effect (resulting in the reproductive isolation needed for maintaining species integrity) is something the entire organism has to adapt to. Having adapted, it seems not unlikely that, in some cases, an artificial changing of GC% (as in the Kelkar paper) would be deleterious. This would be particularly evident in the case of 'lower' species that had not superimposed other mechanisms for maintaining reproductive isolation. Absence of superimposed mechanisms would prevent GC% values from seeking new equilibrium positions. For more see my text Evolutionary Bioinformatics (Springer, New York, 2011).

On 2015 Apr 26, Donald Forsdyke commented:

The data in this interesting paper seem not incompatible with the hypothesis that the need to prevent recombination with other organisms drives a organisms GC%. This anti-recombination selective effect (resulting in the reproductive isolation needed for maintaining species integrity) is something the entire organism has to adapt to. Having adapted, it seems not unlikely that, in some cases, an artificial changing of GC% (as in the Kelkar paper) would be deleterious. This would be particularly evident in the case of 'lower' species that had not superimposed other mechanisms for maintaining reproductive isolation. Absence of superimposed mechanisms would prevent GC% values from seeking new equilibrium positions. For more see my text Evolutionary Bioinformatics (Springer, New York, 2011).