Very good points here.
I failed to remember that captchas are indeed used for other things besides registering as a member to a site. Things like one-time viewing of information (eg: WHOIS), etc, will probably benefit a lot from this. I showed some oversight claiming this service was dumb. It's great for things where it's OK that I get insulted, because I want to see something bad enough anyway.
I was in the mindset of imagining this being on a registration form to become a user of a website. I think the money lost from the amount of users getting turned off by this would be greater than the one-time profit incurred whenever a user registers. That is, unless your site profits from less users. In equation form:
(amt profit per lifetime of user) x (number of users that won't sign up because of this) > (amt of users that do sign up with this) x N [where N is how much you make from this ad captcha].
Notice that the longer you plan on retaining users, the less you should be willing to risk slowing down your sign-ups, unless this ad captcha offers a high enough profit. It is my opinion that for sites seeking long-term relationships with users, that this thing sucks. On the other hand, if you can afford that the user not continue beyond a point, then it's great. That's why it'll work for porn and other seedy crappy sites, and probably why I have low regard for it. I admit this is a foolish mindset.
In regards to "actual problem of spam," I was referring to the answer of the captcha being in the DOM, and the otherwise flawed implementation of it.