That page hasn't aged all that well. The prediction that applications would need to switch to a new hash function "every few years" hasn't panned out. The feared improved attacks on SHA-2 have failed to materialize. Applications that chose SHA-2 20 years ago are still quite secure today.

It seems we just weren't very good at designing hash functions in the 90s.

"In 2007, the NIST launched the SHA-3 competition" and the following year, 2008 SHA-2 is labeled "Minor weakness".

Well, great timing on that competition!