It appears to me that the "horse hedge" article is a poorly written attempt to propose generating profits from the parimutuel pools' inefficiencies, and take advantage of the tendency for players to overbet longshots.
Of course the data presented simply confirms the issues involved (handicapping to true odds/chances to win, track takeout, late changing odds, covering many combinations) make it difficult to achieve a profitable efficiency.
Bottom line - you lose less betting chalk. Well, duh.