Hmmm… More quick results have bonuses applied and yet quick ratings tend to be lower. That would tend to add weight to the theory that a greater proportion of rapidly improving players has caused quick ratings to be deflated relative to regular ratings. (2.9% may not seem like much of a difference, but remember that players are getting that many more bonuses while playing opponents that may be up to 100 points under-rated – relative to the regular rating system)
This would also explain why the difference is greater for those that have had their quick ratings longer. The initial quick rating (set from the regular rating) will decline from playing under-rated players. The longer you play quick games, the bigger the difference should be.
IF that turns out to be part of the difference, then (as long as quick games have a higher proportion of scholastic and other rapidly improving players) you’ll have to add some sort of inflationary “fudge factor” to quick ratings to keep the two systems in balance.