"What" is straightforward: L4 and L5 are two points on an orbit that are 60 degrees ahead and 60 degrees behind the planet, and so move as the planet orbits. If the planet is less than about 1/25 of the mass of the central object, then a combination of centrifugal and Coriolis forces will cause the L4 and L5 points to be places where a third body can have a stable orbit.
Asteroids that have orbits at (or near) the L4 and L5 points are known as "Trojans" Jupiter has hundreds of Trojans that share its orbit, even Earth has one 2010 TK7 (and perhaps more).
"Why" is harder. It seems that this is chosen to ensure that the "planet" is significantly smaller in mass than the central "star". So that we can clearly claim that "the planet orbits the star" and not "there are two roughly equal-sized bodies in orbit around each other". Such configurations would be a binary system and not a planet in orbit.
The significance would be in the nomenclature. In a binary (or multiple) system, the objects are named "on the same level" (for example Alpha-Cen A, Alpha Cen B and Alpha Cen C) Planets, however, are named at a lower level, with lower case letters (Alpha Cen Cb) There isn't really another reason for specifying an upper limit to the size of second body.
It also seems to be a proxy for the mode of formation (this is hinted at by the parenthetical comment). Two roughly equal bodies will have formed by gravitational collapse of a cloud of gas that split in two as it collapsed to form two objects mutually orbiting each other. Whereas classical "planets" form in the disc surrounding a proto-star as a secondary process. But rather than use the unobservable "mode of formation" as a definition, the IAU (in its infinte wisdom) chooses the astronomically meaningful L4/L5 instability