It is always the case that if a test taker has quant and verbal percentiles that are close to each other like yours are the overall percentage will be higher than either of the sub sections.
This is because most test takers find that either the verbal or the quant is significantly weaker than the other. You could say that consistency is rewarded - but so is excelling in one phase or the other. As mentioned by a few people above, verbal scores are, on average, lower than quant scores and there is more room at the top end of the verbal to really excel so a verbal score can get your to 700 even with an average day on the quant - but one cannot get a really high score with one section alone.
As to the mechanics of the scaled scores - they work this way, the computer figures out how you compare to others based on whether you answered all of the questions, the number that you got right, and the difficulty of those questions. The scaled scores just correspond to the percentage of people that the computer feels you are above and below. Please also keep in mind those insidious "experimental questions" that do not count. If you get lucky and lots of the questions you missed don't count that can help your score -- if you get unlucky and happen to ace the experimentals and all the ones that you missed count against you -- well that is why the GMAT is not purely skill based but does have an element of chance...
