Google has become rich for using Proxy Variables as a means to derive inferences. Whether or not its true, if your proxy-sample is representative of the population at large, you still get a valid sampling.
That Google wouldn't know about unconscious bias or how "objective
factors" can be correlated to physical sex sounds ludicrous.
While women taking off for having babies is cited as a reason for pay inequity, Google has provided on-site day-care and facilities to lower the effect of those problems.
Part of the problem is that while men may be given equal time off to bond & spend time with newborns, they may not take it as often. Should that be a positive trait that is rewarded? I.e. if someone works an extra 20-30 hours/week over a "40 hour week", should they be rewarded for the long term effects their work-life imbalance, statistically, is likely to cause.
In the case of professional football players, society says "yes" -- they get paid ludicrous sums in their prime so they don't have to make money past their 30's.
One begins to see the reason for having a 40-hour work week and requirements for 50-100% pay to be applied to work over 40-hours. Why shouldn't software engineers be on a similar pay-schedule? Then if it is the case that men earn more because they work more, it will be clear as to why.
As long as it is presumed men and women work @ similar "40-hour-a-week" jobs, then pay inequities can't be justified to the extent they exist, but for them to be documented, "professional positions" would need to have weekly hours documented like any "hourly" employee.
How much of the extra "productivity" that men produce is due to them being willing to give up any non-work life while they are younger -- with industry responding by discriminating against older workers who start to have families and realize that a life that is filled only with work isn't so fulfilling as they get older?