Turns out most really don’t perform all that well, and that URL shorteners actually increase the load time of pages significantly. As you can tell from the graph embedded above, a lot of URL shortening services add half to nearly a full second to page load times.I don't see how such delays make "a world of difference." Economists sometimes do zany things like multiplying each 1-second delay by the billions of times short URLs are clicked a day and then by the average wage rate to calculate some omnious figure representing the loss to society caused by slow URL shortening services.
To measure this, WatchMouse checked each URL shortener every five minutes from one of its monitoring stations, which are located across the globe. For each short URL, only the redirection was measured, not the actual loading of the target page.
Pingdom did similar research on the speed and reliability of URL shortening services in August 2009, although they only looked at independent URL shorteners and not the ones from Microsoft, Facebook and Google.
Google does a pretty good job in terms of performance with Goo.gl and YouTu.be, but it still takes those about 1/3 of a second to resolve pages, which makes a world of difference if you think about how many website addresses get shortened on a daily basis.
On an individual level, a short delay isn't much of a hassle. If you click a short URL, you aren't that much better off if it loads 1 second faster. To argue that I'm doing something like 7 seconds less of work a day and calling this a real impact in the aggregate is foolish. Even if you open dozens of short URLs in separate tabs, you can't read all of them at once anyway; the rest will load while you're reading the first few. Perhaps this delay will impact your local servers, but only marginally so. We certainly wouldn't be millions of dollars better off if such delays didn't exist.