Red or Gold, the key thing is to mount the sender such that flow into and out of the cube is smooth and stable. As Mike mentioned, mounting anything immediately upstream will cause erratic readings. I've seen bunches of problems caused by fuel filters, 90 deg elbows, or other stuff being too close to the cube. The instructions state to route the fuel line such that vapor can't be trapped within the cube. This can be accomplished with either a slope up or down of the line, but you don't want it dead flat, and definitely not at the top of a curve.
Also worth noting is that the "K-Factor" is different for each cube due to their different internal construction. I use the Red Cube with my Dynon Skyview and the pre-programmed Red Cube K-Factor has been dead on. I'm not sure what the factor is, but apparently Dynon knows their stuff! I didn't intend to buy a Red Cube, but that's what came with my Skyview, so that's what I used.
Looking at the specs of both cubes (link below to specs), the Gold Cube has less resistance to turning (e.g. less pressure drop), so therefor is more suitable to high-flow systems (up to 125 gph flow rate). However, this low-resistance would magnify the effects of turbulence into the unit and show itself as fluctuating flow readings, which is what Mike experienced. The Red Cube has more internal resistance and a higher pressure drop, but the readings are likely more stable. The pressure drop could be a problem for fuel systems that deliver marginal flow to begin with, so take that into account before going this route. My fuel system tests at 15+ gph through the system as installed (including Red Cube), so it can be made to work.
https://www.emapa.aero/Electronics-Inte ... sducer.htmJeff