The U.S. is producing more 2:15 marathoners than it did circa Y2K
...and probably about the same number it did in the 1980s, adjusting for population growth since then.
I "somehow" came across Track and Field News' top-50 U.S. men's performance lists from 2001. Post-Olympic years are sometimes lackluster in this area, but plus or minus years ago, this was unquestionably part of a nadir in American distance running.
For the year, there were only seven sub-13:30s and two sub-28:00s, and a high-schooler had the fastest mile time. Not one man ran a marathon at under 5:00 pace, only two had a faster performance than Paula Radcliffe's soon-to-be-run (and still-standing) WR, only 19 broke 2:20:00, and only 48 ran under 2:25:00. And there's no way I should have been the seventh American at the Boston Marathon and 41st on the U.S. list for the year with a time that would virtually never win the women's division of that race in similar weather conditions.
Using 2017 as a gauge, track distance events have not really become much deeper if one adopts the debatable tactic of factoring out naturalized Americans to assess "intrinsic" talent. Last year, 17 American men broke 13:30, eight of whom were born in Africa. Six broke 28:00, but no U.S.-born runner did.
The marathon is a different story. Last year, 35 men broke 2:19:00 (the current Olympic Trials and Olympic Marathon standard) on loop courses, and 40 more did so on courses considered aided (three at Boston, three at St. George and 34 at CIM/USATF Champs alone; St. George is considered too aided to be used for Trials or Olympics qualification purposes).
The take-home message? I was bored enough to Google mentions of my own fastest marathon and vomit up a passing response. This is it.