How much extra could pool hoppers make? 50%? I guess it required many proportional pools to hop between though.
Meni Rosenfeld derived an equation which approximates this (Table B1 in https://bitcoil.co.il/pool_analysis.pdf):
E(no fallback) = m*log(m)/(m-1)
E = expected share value multiplier, m = number of proportional pools.
To get 1.5x (as you suggested) you'd need three prop pools if you weren't using any fair pools in between prop rounds (no fallback). If you use a fair pool in between prop rounds, then you expect better than 1.5 from only two prop pools.
And I guess Slush's pool is much less profitable to pool hop than a prop pool.
As ck wrote, I think Slush's is no longer hoppable, some kind of PPLNS with an exponentially decaying n.
OK folks, we have a prop pool listed for the first time in a long time.
* abracadabra fires up bithopper
It's good to work out those GPU's from time to time
Couldn't you use an ASIC with bitHopper?