Access to acute MRI is a big hurdle. Having an MRI machine is one thing -- being able to immediately clear it for each and every acute stroke, with easy access from the ED even for unstable patients, is a real luxury.
Ultimately, I agree with you that patients' physiology probably means a lot more than some number on a clock, but if that clock is usually a good approximation, then it will probably remain in wider use. The MR-WITNESS study is one of the U.S. based trials looking at this. It really comes down to whether you want to personalize therapy, or sacrifice some of that granularity to make your treatment more widely applicable. Kind of a biology vs. public health debate. Maybe this sort of research will drive imaging technology toward smaller and cheaper imaging surrogates.