The pipe bends and length do matter, but they aren't solely what determines dynamic head, since static head is a bigger part of that. No matter how the water makes its way into the barrel, static head is set by the height of the water in the barrel, not the pipe routing. There's probably a better way of explaining it but I'm tired...
My post said one route has more head than the other. I don't think I ever said the extra bends and length were solely what
determines all the dynamic head, or whether dynamic or static head makes up the majority of head. But if one has more pipe
and more 90 degree fittings (the up and over version) than the other (fed from the bottom of the filter version), and they both
have basically the same static head, then I guess I'm missing how one doesn't have more total head than the other. I'd be
interested to hear how that works when you have time. I'm here to learn.