In the past a lot of discussion, plenty of ideas. Even in our group building/flying the LEU Treehugger and DE Stubby there were lively discussions. Here is what we did.
LEU Treehugger has one tank mounted in the left wing. It had one outlet at the rear bottom surface. The filler port on the top was vented with a tube pointed into the airflow. We used 3/8" urethane and aluminum tubing from the tank outlet to the "T" near the carburetors. The run was sloped down for its length to allow air bubbles to percolate up. The relatively large size of the tubing resulted in the fuel flow velocity through the tubing to be relatively slow, allowing the bubbles to go up at the same time as fuel going down to the engine. Treehugger had its problems, but never had any fuel flow issues after we went to the alum tank and this arrangement. We had a strainer in the tank outlet and a valve there, too. I've attached a pic showing the line.
DE Stubby has two tanks, one in right wing, one in the left. We went round and round about the fuel system, considering a single fuel selector valve, two separate valves, one common valve, one or two outlets in each tank, on and on. We finally ended up with two outlets, one in the side wall at the rear near the bottom, one in the side wall at the front near the bottom. This was to answer the concern about losing fuel if in a nose down attitude. The individual tank outlets were joined together, then those two lines came together in a "mini header tank", and fuel from one tank can equalize to the other through this, and it gives a couple minutes of run time if there are any issues. From that a single valve was installed, then a 3/8" line to the gascolator at the bottom of the firewall. The mini tank has a vent at its top, and a line from it to a "T" near the top of the wing surface. This urethane tube serves as the visual fuel level gauge. The sides of the "T" go to the tank vents, located on the side wall near the top. The filler nozzle has the vent tube facing into the airflow. Complicated as heck, even harder to pipe it all up. But, we haven't found any possible fault in it during operation. One potential long-term fault is that the fuel valve is located downstream from the urethane vent/level gauge line, and if the tube would fail when parked, a lot of fuel could leak onto the floor. So, we keep an eye on the urethane lines. I've attached another pic showing some of that system.
I'm not saying these ideas are the best or only correct ways, but they did and are working for us.
