When do you need to take a year off to do research? What should your research profile look like come ERAS time? Obviously an M3 going in with 0 prior research would benefit from taking a year off, but how about someone with a few NSGY-related publications under their belt? When does a year off for research start creating diminishing returns?
You need a step score first. Under a 240 and there is no point. All you really need is a high step score to match.
What do you mean under 240 there is no point? I imagine those would be the applicants who could benefit from a research year.
This match has shown that Step 1 doesnt matter as much as the intangibles.
Dude, that is an isolated case. 1 (or ok 2 ) matches out of 200+ is not a trend, they were merely outliers. Your average white/yellow/brown dude ain't matching w/o beastly scores.
Regardless, your average applicant is judged first and foremost by that one score. It's always the first question your chair asks you.
To actually answer the question (since other posters seem to prefer just jerk each other off about high scores), I think post-M2 is good to do if you can get into something like the NIH Fellowship, and then pending whether you're still behind the average pub number (I think it's 16 as of 15-16, but people make little shitty posters/abstracts and those count in the Resident Report), you should strongly consider it after M3.
Yes, the scores matter. But you like anyone else can go to the data and see that mean Step 1 = 247 with a large SD (13), so there's that spread for a reason (i.e., significant research experience, being a NFL player, etc).