I do not think the term put-call gamma imbalance refers to the imbalance between "a put" and "a call" at the same strike. . .so put call parity lives on.
As for the exacerbated volatility, my experience has shown that there is pressure to "pin" at a strike which would decrease volatility near expiry.
scenario:
Strike = 100
Stock =100
expiry in 10 minutes
all else equal
Those traders long the 100 straddle will scalp their gamma aggressively in the final moments. As the stock trades below 100, the 100 straddle will generate short stock for the long holder. The long option holder buys stock as it drops and sells it as it rallies butting pressure on the stock to stay at 100.
The short option holder locks in losses every time he hedges so is more incline to not take such energetic hedging action as that of the longs.
Now, if the market makes a surprise move of substance. . .all bets are off, there may exist a very large delta imbalance (short option holders) and they might be forced (pain of large loses changes everything) to sell their long deltas on the downside or buy in their short delta on the upside. In either case this will increase volatility . . .until the next trike is reached and the "pinning" battle can begin again.
this article gives a better explanation than mine :-)
http://realmoney.thestreet.com/articles/12/15/2011/how-options-expiration-affects-stock-prices