Say what? Everyone from old ladies to the super elite run nearly the same speed? Not exactly true, but the idea is that the difference isn't all that big, either.
For instance, what time does a fast runner complete a 10K in? About 30 minutes. And a slow person? About an hour. Therefore most runners complete a distance at speed v or 2v (where v here equals 2.8 m/s). People do take longer than an hour to run a 10k, but they're usually walking for parts of it. More importantly people who are racing (not just running, i.e. trained and pushed themselves fairly hard) will run under an hour.
Earlier I linked a graph for energy cost of running versus walking. I didn't talk about walking in that post, but what the graph shows is that below 2 m/s walking is more economical than running. That speed translates into a 10k time of one hour and 23 minutes. Sure enough among the 8393 runners in the 2011 Ottawa 10k about 7800 of whom (92%) ran this chip time or faster. There were plenty of walkers to account for the remainder. If we wanted to be generous we could say then nearly ever human on the planet will run a particular distance D between speed v and 3v (or alternatively between time t and 3t).
As a physical chemist I'd be ecstatic if I knew rate constants with that precision. I'm often happy with a order of magnitude precision. But with running, before I've seen you I can be fairly certain that you'll run a marathon between 2.5 and 5 hours. A single look and it's easy to know your time with twice that precision, probably more.
Then I imagine the huge range of speeds we experience or observe in a lifetime. An expert cyclist can cover 42 kilometers in under an hour. Anyone can drive 42 kilometers in 21 minutes (at 120 km/h). Cruising airplanes cover the marathon distance in about four minutes. At the other extreme, a snail covers the same distance in half a year. That I can safely predict you'll run in such a narrow time constraint as t and 2t is pretty remarkable. But it gets better.
If I know how fast you already ran a race, predicting future performances is a snap; I can make very accurate time predictions. For an individual we tend to exaggerate our time improvements over the course of a career. For instance, when I was 19 I could run 10 kilometers in 36 minutes (with no specific training). After a few years of proper training I could run it in 32 minutes, or 12.5% faster. Huge improvement? It doesn't sound that big when phrased that way. Yet if I gained another 12.5% improvement in speed I'd be world-class; a 28:27 guy is flying, probably sponsored, and certain to win any chosen local race. In a way these are huge improvements.
Take a closer look and the changes further shrink; within a single season I might hope to better my time from, say, 32 to 31 minutes. This is a 3% drop in time. Does any 32 minute 10k runner think that running a minute faster (which seems like a lot) is only 3% faster than he's already going? I doubt it. Likewise a 2:30 marathoner probably dreams of running 2:20 (at least I do, sometimes), but even that's only 7% faster! Objectively speaking the differences look to be almost within the margins of a calculation error. By comparison your body weight fluctuates by more than 7% within a marathon. Seen one way the time differences are tiny; seen another way they're huge. A slight paradox, to be sure.
I smile when a study reports something like "Our experiment used a group of N elite runners who ran, on average, 32.3 +/- 1.5 minutes" (I'm making these specific numbers up, but they're typical). Not surprisingly, no sub-31 minute guy would confuse himself with an over-33 minute runner, or vice-versa. But to scientists they're about the same speed, or as close as they can gather for their study. Combining results of apparently different individuals who on the surface seem clustered is tricky territory. In my own experience I find runners don't consider themselves of 'equal' ability to another until the time differences are about 2% (i.e. 40 seconds for a 32 minute 10k-er, 5 seconds for a 4-min 1500m runner and 3 minutes for a 2:30 marathoner).
Do we really, then, all run the same speed? It depends who you ask. Personally I consider viewing the differences as small to be a great motivator. Hey, I'm only 10% away from making the podium in some big races. We're hung up thinking of ways in gaining that extra second or two when really it's not that much; A minor improvement in your physiology produces minor gains in speed, but that's all you need. Paradox resolved? Not sure. No matter the answer, because the improvements will be minor, i.e.the changes in your body subtle, the body ultimately will control them. In other words your conscious mind isn't a big player; you can't just 'push' harder to win. What to do? No simple solutions, oh well, but at the same time we're all inches from doing incredible stuff. Fascinating.