So what i was thinking about when reading this thread is a bouncy ball. A bouncy ball returns most of the energy you put into it when it flexes. If you took a ball of steel and dropped it it would not return as much of the energy. Does this mean that rubber is a better biking material? Obviously not but it points out that perhaps there can be 2 stiff bikes, 1 that returns the energy and another that doesn't as well.

So can anyone who understands this better explain how maybe a bike could be flexy but in an efficient way? It makes sense that stiffness is overall better if you consider this scenario:

Imagine you are pedaling on your bike and therefore the chain stay is being compressed. The chain stay shrinks a small amount and wants to rebound back into its regular size/shape. As you finish your stroke and lead into the deader areas at the top and bottom of stroke you could slack the chain letting the chain stay expand again without transferring energy into the drive train.

If you kept up the tension evenly through the stroke i would imagine the chain stay would reach a point that it wont flex anymore and will be transferring almost 100% of your power into the cassette. Or as you slack a little on the stroke the chain stay would expand a bit transfering the stored power (minus the loss based on the material) back into the cassette.

That seems to make logical sense to me. Is there anything wrong with that view of whats going on in a bike?