Hence, we time things. We have performance depth charts to see where we stand against both our contemporaries and our predecessors and to set qualifying marks for the more exclusive meets. And we also have those "silly" conversion charts which attempt, albeit imperfectly, to measure the quality of a performance against those achieved in different events. It's in our nature to question whether Steve Scott's mile record is actually a better performance than Jim Ryun's former record run on a cinder track. We also wonder if Muhammed Ali was a better boxer than Mike Tyson, Evander Holyfield or Lennox Lewis. In boxing, the question of who's the all-time best is virtually pure speculation. In track and field, we have objective measurements available to help us make our speculation a bit more scientific. There are some excellent statistical methods for comparing performances in different events, as well as methods which give a pretty good idea of how much people tend to slow down as the race distances get longer. And enough people have run on both cinder and all-weather tracks that the guesswork in comparing times on the two surfaces can at least have some reasonable ballpark numbers to back it up.
A conversion chart is just an attempt to provide a kind of Rosetta stone to "translate" performances. Certainly you can determine how you stack up aganist the competition in several different events by racing people in all those events or looking at your position on a depth chart, but what does that tell us about events that are less-contested? Comparing time-for-place in races or using all-time lists doesn't tell the whole story as to how strong a one hour half marathon is compared to a 13:00 5,000. There have been many paced 5,000s, while the half marathon is a relatively new event as a global championship and the best runners in the world haven't tackled the event nearly as often, so we can't tell if the half record is weak simply by looking at the all-time lists. But we can use deep lists for the more widely-contested traditional events from the mile through the marathon and can determine on average how much runners tend to slow down as the races get longer. Since many races of the same distance are run on both the track and the road, we can also use statistics to make reasonable adjustments for track vs. road. And this gives us a more numerical insight into comparing performances in different events. It stands to reason this is at least somewhat more accurate than saying the world record for 30k on the track is comparable to the 10,000 record. We all know it isn't, but just how weak is that record? Having a more reliable basis for comparison might prove useful when determining qualifying standards for meets or when assigning point values to performances, as in the decathlon. "Age-graded" performances attempt to do the same thing.
So think of it as "silly" if you like, but some people are interested in conversions, perhaps to come up with a ballpark time or pace to shoot for in an unfamiliar race distance, to see which events really are their best ones, to assess where they stand in some kind of local or regional "performance points" seasonal competition, to choose which race they should run (or try to qualify for) at an important outing, etc. You can't always tell these things by racing other people.
And since today's runners can't race those of other eras when everybody was in peak form, we use the stopwatch as one way of comparing eras. It isn't perfect, of course. Track surfaces are faster, shoes are better designed, travel is easier, rabbits are plentiful, competition is deeper and money is on the table (and maybe some drugs). But the stopwatch tells us there can't be much doubt that Bekele is running at a much higher level than Zatopek or even Viren or that numerous high school runners today can run faster than Nurmi ever did. And records at any level (personal, school, meet, state, national, world) do serve as tangible challenges for athletes of all abilities - like you said, to set a new standard.
