If Google was more specific, fraudsters could probably figure out how many fake clicks per day per IP are allowed, etc.. While innocent people like the author will get banned once or twice, fraudsters will probably go through tons of accounts, maybe one for each family member, for example. So the fraudsters are going to be picking up data due to experiencing many more bans than innocents.
Heck, if the innocents all blog about it afterwards like this one, they are just acting like an extra data point for the fraudsters anyway, and most of them have so little traffic that banning them or not and telling them details or not will never make a difference to Google. So hiding details from them hides them from fraudsters too, although treating them this way is probably due to the expense of dedicating resources to worthless customers more than that. Let's say you have a million users and 10 customer support reps. Which users are you going to support and tailor your polices to? The ones with appreciable traffic who make most of your money.