The gs3 is a partial differential design.
A vexing problem solved is the changes in gas chemical composition and the resulting changes in thermal conductivity between the active and reference channels. By sharing the same container along the tube, the only gaseous conductivity differences would stem from diffusion effects. Imagine one end of the system generating gas phase chemical changes. I think these concentration differences would rapidly vanish as high temperature thermal diffusion drives both sides of the system to near equilibrium.
Are there any other differences between thermal conductivity in the active and reference channels? Obviously the active channel is filled with highly thermally conductive metal powder while the reference side is filled with low conductivity gaseous hydrogen related moieties. For purposes of simplifying the modeling the thermal system, let us assume the center of the reactor core is the origin of heat transport. Which half of of differential system is going to have the most thermal conductivity to the outside world? I will guess that the powdered metal side of the reactor will have the most thermal conductivity and hence the outside temperature will be highest on the metal filled side. This positive temperature increase will be caused by differences in the thermal conductivity pathways (one side is powdered/sintered metal verses the hydrogen/allotrope gaseous transfer medium). Without hard conductivity numbers for the various substances but working under intuition, I would assume the metal is a better conductor of heat than almost all gases. Therefore the outer surface temperature should naturally be higher in the "active" zone (without LENR) and the claims for excess heat should be very cautiously considered.

jdk 03jun2015 22:25 central daylight time

Hi jdk

the GS4 calibration alleviated these fears.

In the end - only statistically significant elemental and/or isotopic shifts provide hard evidence where some sort of flow calorimetry is not used.