Was thinking about it the other day, seemed to show you can "Google Bomb" for a competitive term, and the "Google Bowl" theory of a penalty for too much similar link text is most likely a myth, or at least needs to be qualified. Apart from numbers of links, I suspect the positive results were due to the range and diversity of the linking sites, something hard to "manufacture", and the pattern must have followed quite closely a "genuine" GoogleBomb one. Is there a thread that does a PM on the experiment?
I am also suspicious of both the penalization idea from not varying your link text or too many inbound links in a given period of time. Because there are definite esxamples of websites that are not penalized by these factors (websites that have an article or press release get posted on thousands of websites with similar or identical link text), it is hard to confirm that penalties really get applied for this kind of thing. There are of course people who feel that they have been penalized for these things, but as with all other things SEO it is difficult to proove exactly what the cause was.