I disgree with that.
Handicaps are determined by taking the ten lowest of the last twenty scores, thus, a handicap is skewed toward each golfers "better" game. If you took the lowest 20 of the last 20 handicaps would jump. A lower handicap has narrower margins, thus, he's at an advantage when competing against higher handicaps, especially if conditions are unusual.
That's true when you are playing match play, and only playing against a single opponent. But if a low handicap player plays in a handicapped stroke play event statistics dictates he is very unlikely to win. The USGA system was designed to equalize match play, not stroke play tournaments with large fields with players of different skills (which is why many such tournaments are flighted and played at scratch) The problem is that the standard deviation of the many guys with the high handicaps will overcome the low handicap's lower standard deviation and totally swamp the "bonus for excellence" built into the USGA system.
The USGA publishes this neat little table with the odds for exceptional tournament scores for various handicaps, which is calculated based on the math Dean Knuth used to develop the USGA handicap system. As anyone with a USGA handicap knows, it is tough to shoot better than your handicap -- its 5:1 against just to match it over the whole handicap range. But it quickly diverges as look at beating your handicap. Beating your handicap by 5 strokes is 379:1 against for a 0-5 handicap, but only 72:1 for a 22-30 handicap. The odds are slightly better, at 359:1, for a 30+ handicapper to beat his handicap by NINE strokes!
Even if no one sandbags, the low handicappers really have the odds stacked against them in a handicapped stroke play event. There will always be some 100 shooter who has one of those "euruka" moments on the first tee and figures out how to cure his slice for one golden day and shoots a 91, and there's no way a 5 handicap is going to shoot a 68 to match him.
http://www.usga.org/playing/handicaps/understanding_handicap/articles/deanstable.html