Heck, I'd bet an algorithm would've caught it (or at least flagged it, caused a review, and had it confirmed) a loooong time ago. Which would've cut off his income while it was smaller. A smooth-functioning, impersonal algorithm may have been preferable in this case.
Google could provide a kind of back-and-forth here, to restore the ability of a site author to provide a testimonial for a particular advertiser. An author would have to provide a (short and limited) list of sites and testimonial copy for each one.
A sailing site could then, for a particular sailcloth manufacturer's ad, provide a short testimonial noting the high quality of the product, etc...and Google could serve up that custom testimonial along with the ad.
You'd either have happy advertisers, or not.
There's a difference between "I support this product and you should check them out" and "click on these ads." Google's ads are targeted and beyond them being about sailing, he doesn't necessarily know what's being shown.
So it's pretty difficult for him to personally recommend particular ads or companies. I'd like to see the exact messaging he sent to his subscribers.
Google correctly identified his business as dishonest and terminated his account per agreement that the guy signed when he started using AdSense.
You try to make a case for Google trying to determine shades of gray, but it's a non-scalable, slippery slope.
The current rule is crystal clear and simple: don't encourage users to click on ads they otherwise wouldn't. By definition every such click is fraudulent because the user wouldn't otherwise perform it.
Google (or anyone else, for that matter) cannot implement more subtle rules (that weight factors like how many times someone did it, how often does he do it, how subtle the encouragement is, how many kids he has to support) in a scalable way.
Google couldn't even apply more subtle rules in a consistent way, giving fraudsters even more weapon in what is a PR game (take all the ample, emotional justification from the above post AND allow that fraudster to point to other people who are doing similar things but Google hasn't taken action against them because some fallible person applied a different subtle standard for what fraud is than some other fallible person).
As this thread demonstrates, there's plenty of people willing to sympathise with the fraudster despite clear, self-admitted evidence of wrongdoing based purely on his self-professed (hence extremely biased) version of how good of a person he is and how good intentions were and how badly he was treated by big, cold Google. Emotion can trump facts.