[jessriedel]

> Remember, most of the time the LHC isn't actually colliding,

Huh? I'm pretty sure the LHC is designed to spend >50% of it's time colliding. Lately they've had to spend more downtime upgrading the machine because it's been finicky, but this if definitely hurting the project. The decisions about what risks to take and how much downtime to allow are hugely contested.

> The bottle-neck for LHC isn't how many collisions it can throw up...

The number of collisions is definitely a bottleneck. Yes, to some extent you can make up for a small data set by throwing more grad-students at the data to create a more clever analysis, but that really doesn't take you far. Searches for most major new physics scenarios have very well-defined minimum amounts of data. If the requisite amount of data isn'y taken in the LHC's 1-2 decade lifetime (because of delays/budget/whatever), and another accelerator isn't built, then we simply won't be able to evaluate that scenario.