This plan seems to be far more valuable for definite affiliate offers. The part about " hopefully Google will display Ads that are costing others the $30 per click" seems to be giving up the control be are trying to build. Although it does get around the issue of AdSense not providing low value clicks, adding another step in the funnel will have its own conversion rate so guessing isn't a good method IMHO.
On the other hand, the idea of having a more generic landing page (centered around an ad group for example) that has satellite sub-niche landing pages to convert the lower volume (higher cost) but more targeted traffic seems valuable. It makes the generic landing page a static niche screener that funnels traffic to the right sub-niches to convert better through better matching traffic with more focused offers.
Remember that there will be an additional conversion rate as well so the spread between the value of the landing page's (lp) initial traffic and the targeted satellite (sat) traffic will have to be higher (e.g. $0.10 -> $1.00 for a 10x gain vs $0.10 x 10 = $1.00 (lp to sat 1/10th traffic), then it would effectively be $2.00 -> $20.00 for the same 10x more targeted conversion).
If AdSense will add another "conversion rate" from auto-picking the ad, then that last 10x will depend on their targeting more than ours and may reduce the value of the technique. But then again that is always there with AdSense
Ad Dennis Miller likes to say, "But then again I may be wrong!"
I'd love for Marcus to add to this discussion.