The article you're referencing is 5 years old. While the Tor network is still fairly slow, I believe it's come a long way since then.
How do we define "slow"?
Either by quantifying the speed, or by reference to something most people are more familiar with. Jobiwan's answer does the latter, using a comparison to an HTTPS connection. Qualifiers like "very" or "really" are relative and subjective, so for the article to say it's "veeeeery" slow doesn't help unless you know the baseline from which the author is commenting.
If it's numbers you want, have a look at the Tor Metrics pages. This gives an ongoing, historical record of download times for files of different test sizes. It should provide some idea of network latency. That said, there are no "Tor vs. the clearweb" comparison charts.
For details on why Tor is "slow" (yes, bouncing traffic around different relays), have a look at this entry in the Tor FAQ.
Is there any way to change the delay?
Probably not without fiddling about with the source code to change the number of hops your traffic goes through.
If you're happy to do that, and if you're unconcerned with the anonymity afforded by Tor - which seems to be the case, given the use-case in the article - then take a look at the following discussion: How can one make Tor faster, when willing to to sacrifice anonymity?
(It's possible some of the suggestions in there would help, though from the comments there seems to be some debate.)
That said, if you're unconcerned about being anonymous, perhaps a VPN or standard proxy would be a solution worth considering.