Obviously 5K track is faster than road (as a rule). But given the extra adrenaline in a race, being pulled by others etc, does a 5K solo time trial on a track approximate pretty closely to what you could race on the road, on a decent flat course?
So if I ran an all-out time trial solo on the track of 20:00, could I reasonably expect to run the same in my local 5K the next week?
I'm wondering also if this is why the advice here is that, for example, 5 x 1K intervals with 1 or 2 minutes rest are a good race predictor?