They're looking to the [what they think is the near] future. Future proofing is generally considered to be a good idea, but I can't help but find a fundamental flaw in making quad-SLi a reality with the current GPU's available. In the time it takes for any game to come even close to utilising that much power, we will probably be into a whole new generation of graphics hardware- there will be new technologies which we will be wanting on our cards somewhere down the line making the current ideas defunct. Couple this with the fact that no game or forseeable future game will require the kind of juice which quad-SLi will yield (2Gb of dedicated video ram!), I think it's nothing more than going way over the top just because they can- bit of a gimic really. Benchmarkers may find a use for it, but I dont realistically see it being a necessary thing.