"do you think TSH over 2 would be sub-optimal for runners? Say someone came in at 2.5, would that have a significant negative effect from an athletic standpoint?"
Hard to know. Medicine is more interested in where to draw the line between disease and no disease, and also the implications of that decision. As the WSJ piece points out, a decision to treat more aggressively would double the number of hypothyroid pts in the US (adding over 15 million pts). Even though the drug is cheap, toss in the office visits and labs and you're talking serious money, in a system that already spends too much.
My wife was put on synthroid when her TSH approached 10, and then with pregnancy the levels fluctuated a bit, so she bounced around. She says she felt better running when it was in the 2 range, and more tired when it was in the 5 range. But, there was also less sleep with kids, job stuff, not training as much, etc. not to mention the potential placebo effect of thinking you're taking something that will boost your energy. To truly answer that question you'd want to randomize elite athletes to dosing levels targeting different TSH levels, have them train identically, and have they and their coaches be blinded to what their TSH was (including things like sham dose increases in the less-aggressive group). Not an easy study to do, and not sure who would fund it. Any unblinded study is of dubious value, since just the process of being monitored, receiving extra attention, and all the other things that make placebos have an effect come into play.
Finally, say you found that aggressive dosing improved performance by 2%. A big deal to an elite runner. But do you change the definition of hypothyroid based on that, and treat an extra 15 million Americans? Or just treat the ones who want to run a bit faster? What if the guy who just wants to break 20 min in a 5k hears about it, do you treat him? What else do we "optimize"? Testosterone? Growth hormone? Cortisol?