I'm near wit's end trying to find out why I get bubbles in the line from my facet pump to my carburators.

To recap my setup:
Fuselage fuel tank -> fuel selector valve -> fluel flow sensor -> G3 automotive paper filter -> to drain point -> T-split  to each fuel pump. Separate lines from each fuel pump to carbs. Three carbs. The two outer carbs (sharing one pump) each feed a secondary port. The center carb, with its own pump, feeds the two primary ports. One throttle to the outer carbs, one throttle to the center carb. When have climbed above pattern altitude I turn off the pump to the center carb, ready to reduce center carb throttle when all fuel has been burned out of the bowl, and use center carb throttle to lean mixture. Output from pumps near fuselage bottom slopes up to carbs which are appox level with center of fuselage tank.

Bottom of fuselage tank about 16 inches above the facet fuel pumps.
When removing the quick-drain valve, the system gravity feeds about 17-19 GPH through the selector valve, flow sensor and filter.

Various observations:
Idle fuel pressure on ground about 3.8 - 4 psi.
Max power fuel flow near sea level at full power only about 13.5 GPH. 10 GPH at altitude.
Fuel pressure to outer carbs is roughly inversely proportional with observed amount of gasses (air?) in the line from the pump to the outer carbs.
Engine wants to quit when pressure drops to about .1 or .2 psi. (Only happened one time after a continous full power climb to 10,000' to get over clouds. Pressure dropped steadily to that point.)
Removed flow sensor and filter and amount of bubbles appeared to be significantly less.
Put sensor back in, and bubbles reappeared, inconsistently.
Have mostly observed fuel pressure drop at high power climb attitude. Seems to get worse at higher altitudes.
Have checked tank vent (used to be shared between the three tanks, but as part of debugging now goes only to fuselage tank). Vent is pointed forward, mounted below fuselage, center, just forward of main wing spar.
On the last test flight today, I continued to see bubbles, even after reducing power for several minutes and levelling off. Were at 7,500'. Pressure about 1 psi. After a dive to get down below Class B 6,000' floor, pressure came back up to above 3 psi.
A few minutes after landing I noticed that the fuel line to outer carbs appeared empty. Turning on the pump for a few seconds filled it up. After turning off the pump I noticed that the line emptied out. How can this be? By gravity feed alone, the fuel level should have been a couple of feet up that line. I did not delect any gas leak.
Maybe 10 minutes later, the line stayed full after turning on and off the pump (or was now full - not sure).

This brings me to questions about vapor lock. I actually don't know what that is. Carb bowls are vented into airbox. So how could the fuel in the bowls be under pressure, maybe boiling, and push gasses back down the fuel line to the pump? Unlikely? I do have a SS heat shield between carbs and exhaust headers. Does its ability to reflect heat lessen if it is no longer shiny?

I guess the next step is to replace all rubber lines going to the pumps with transparent ones so I can see what's going on the the lines to the pumps.

But other ideas are certainly welcome. Up till the last observation, I've been trying to locate air leaks in lines to pumps and was ready to replace the brand new facet pump with an old one, in case it somehow is sucking air.
Now I'm mystified.

Finn (please no jokes about wit).