I think the theory behind the new system is to account for the evolution of the game playing environment. For example, if we took Ozzie Smith, with all of his athletic skills, and somehow transported him back to 1912 to play shortstop behind Walter Johnson, he wouldn't field as well as he did in his own era. Instead of artificial turf with its perfect hops, he would be playing on rocky, uneven ground on natural grass, unevenly cut and poorly maintained. Likewise, instead of his Wilson A2000 (or whatever he used), his glove would be the equivalent of a poorly made winter mitten. Instead of a bright shiny new ball every other pitch, he'd be chasing after a soft, misshapened lump that looked like someone wiped their butt with it. The end result would be poorer range and more errors.
By considering not only the relative fielding and range of the actual fielder but also the average fielding and range in the years of the pitcher and batter, the new fielding system seems to be trying to approximate some sort of rough average game playing environment. Thus an above average fielder playing behind a pitcher from a good fielding year and pitching to a batter from a good fielding year benefits. Mix in fielders, hitters, or pitchers from the down fielding years in the 19th century and the fielding/range drops somewhat.