On the contrary, this is the exception that proves the rule. The process exists because we can't expect once-in-a-lifetime works to appear from every candidate.
I think there's a precedence for it too. I believe George Dantzig's doctoral advisor offered to accept the solutions to a homework set as a thesis since it solved major open statistical problems.
For those who don't know, here's some background on Dantzig:
"During his first year as a doctoral student at the University of California-Berkeley, Dantzig arrived late to the class of Jerzy Neyman, one of the great founders of modern statistics. On the blackboard were two problems that Dantzig assumed to be homework.
"A few days later I apologized to Neyman for taking so long to do the homework—the problems seemed harder to do than usual," Dantzig once recalled. It turned out the conundrums, which Dantzig solved, were two famous unsolved problems in statistics."
How so? The "process" of getting PhD is writing up his proof and getting it reviewed by the university, and verifying via oral defense that he knows the proof and didn't just plagiarize it.
On the contrary, this is the exception that proves the rule. The process exists because we can't expect once-in-a-lifetime works to appear from every candidate.