Zephir, I don't know how to explain it better than Lubos Motl (26 April 9:04 GMT) already did. I'll make an attempt.
The equivalence principle (of inertial mass and gravitational mass) in general relativity deals with ponderable quantities of mass in relation -- the symmetry of the mutual attraction between them explains the physics of gravity in a field theory without imposing other parameters, because it is the spacetime field that is physically real.
If one tries to introduce discrete energy exchange (i.e., through massless speed of light particles, bosons) as a physically real parameter, one comes up against the fact that there is no elapsed time between these massless particles, therefore no change in the field; measure in any direction at any observer velocity, and the speed will be the same. All that we know about gravity so far is classical physics alone.
Now you know the two-slit experiment, right? This is the heart of quantum mechanics. The two slit experiment informs us that bosons communicate holistically; i.e., these massless particles make coherent wave patterns from discrete units in statistically predictable ways.
Lubos is saying that Verlinde cannot get a workable field theory that includes mass, because if one makes information a physical parameter, it is wavelike, while the deterministic interactions of classical gravity are particle-like. So we are stuck with the same problems of unification that we always had, plus--Lubos claims--an additional problem of incorporating time into the theory. Time in classical physics is a simple parameter of reversible trajectory; in non-relativistic quantum physics, time has no meaning. So when we start talking about information entropy (whose mathematical model is identical to energy entropy) as physically real, we lose time conservation--a fundamental symmetry principle.
Gravity is not physically real in classical physics. If one wants a field theory in which gravity is identical to physical information and information is dissipative, then gravity has to be physically real. There goes the spacetime field, though, because of the nonreversibility of the time parameter.
I know that Lubos Motl's objections are sound, and must be successfully countered. However, I favor Erik Verlinde's and Ted Jacobson's approach -- why? Because I see the answers in a model that exploits what Hawking & Hartle found 30 years ago -- that imaginary time preserves both time reversibility and time asymmetry in a field theory.
If you're interested, my paper "On breaking the time barrier," is here
Tom