Just to make sure you understand, the tornado that occurred on 31 May 2013 was rated EF3 not because it was actually weaker than previously thought, but because the NWS decided to rate it exclusively based on damage (and thus ignoring all radar based observations of wind speeds in it). While it may have killed "only" four people, tornadoes are also not rated based on the number of injuries/deaths they cause.
The study may have used the environment in which the 24 May 2011 EF5 occurred, but based on the accounts I have seen between the two tornadoes, it seems quite possible, if even likely, that the 2013 tornado was a bigger BAMF than the 2011 tornado. Not that it really matters which environment you use - both easily supported top-end tornado production from supercells.