Originally Posted by Cactus

It's not clear to me that there is a stiffness impact on the location of the stays or from a sloping tube. While I don't know that there is no impact, here are the issues that cause me concern with accepting these assertions.

1) I have yet to see a test proclaiming to prove this thesis where the weights/forces are described as being applied to the frames equally. That is, with the shorter seat tube, if the weight/force is placed at the top of the seat collar, then on the longer seat tube (the frame with a horizontal top tube) the weight/force needs to be applied somewhat below the seat collar (the same distance from the BB). Otherwise, the leverage of the longer seat tube will cause greater deflection. That might at first seem reasonable, but the rider doesn't shrink and doesn't sit closer to the bottom bracket when on a compact frame. So, from what I've seen, the comparisons are apples and oranges.

2) A tube is a more efficient stiffening-agent than a bar. Bend a bar, and there is some material in the middle that is neither stretched or compressed - that material doesn't resist the bending with much effect. In a tube, that middle material is missing. On an equalized weight basis, the tube should better resist bending because all of it material is located where the stretching or compression is maximized - and thus encounters the maximal resistence of the material. In the vertical plane, a bike frame is like a giant tube with the down-tube being stretched and the top-tube being compressed. By effectively bringing them closer together with a compact frame, one is reducing their leverage to resist flexing.

A bit of a corallary to this is that the top tube is the most critical tube in the main triangle regarding it's straightness. Under stretching, a tube naturally straightens out - and then as more force is applied, it suffers not from any initial lack of straightness. This is the case of the down-tube. Under compression (the top-tube), any deviation from straightness causes forces to vector out and further bend the tube. Consequently, any initial bend in the tube will be excacerbated by compression and limit the stiffness of the frame. In most cases, this should only be an issue when a frame is highly stressed, that is at its maximal lightness for its assigned task. But, when you see a frame with a curved top-tube, you should assume that the design has not be optimized for the purposes of a bicycle frame.

Most frame tube produced today are pretty darn straight. Still, a good hand-builder checks his tubes before building to ensure that the frame performs as intended.