Thewhole-surface R-value of the floor with gaps isn't much different from one with no-gaps (it would be slightly higher in an ASTM C 518 type test due to the trapped air films at the gaps), but granted, if it was built without a rosin-paper or similar slip-surface it will convect the warmer air of the cavity into the room at some rate.

But I'm very dubious that rate of convected heat off the joist cavity side of the EPDM is going be high enough to deliver 1.5x the heat of the conductive strips even if the convection were completely unimpeded. The ability of the tubing to deliver that heat to the cavity air just isn't 1.5x as high as the conduction to the wood on the contact side, even if it's 2-3x the surface area (and radiated fraction of the heat at these low delta-Ts and tiny emitting surface is a miniscule fraction of the total.)