(I'd love to see what percentage of patents are rejected and how dumb they have to be before they are rejected).
I have a funny anecdote concerning dumb patent. Back in 2002 I applied for a patent in the area of security of Wi-Fi. I first described my invention in a word document, and sent it to the company lawyer. He then turned it into legalese for the patent application (2nd version). I reviewed and found several errors. Basically I was using terms like "computer A, B, C, etc." and he changed them to "computing device #101, #102, #103, etc.". I have no objection except he did not do a "one for one" substitution of computer B with computing device #102 for example. I spent a lot of time fixing all the mistakes (3rd version). Next thing I knew the lawyer actually applied for the patent using the 2nd version.
Now my original invention could be dumb but the modified version is pure nonsense. Then I followed the objections from the patent reviewer and the replies from the company lawyer. First objection is that my idea was lifted from another application except that the other application was filed 7 months later than mine.
This back and forth went on for 5 more years (which turned dumb and dumber) and finally I got my patent approved. The only change to the nonsensical application is some particular legalese expression that the lawyer uses in about 4 or 5 places in the document is now replaced with a different legalese expression. However, the nonsense remains which tells me that in this case, the patent reviewer does not even bother to read my idea, which does not require special knowledge about Wi-Fi or encryption. It is like an elaborate game of 3 cards monte played with encryption keys.