Thanks to your post, I visited the website to check it out myself. Following are my observations (pros and cons) about wikipatents.com:

Pros:

1. Social community based valuation system - enables reviewers to participate in valuating a patent democratically, thus satisfying Web 2.0's emphasis on web democracy

2. Enables the potential end users of patent ideas to rate the patent. I guess this is the most important aspect of wikipatents.com

3. Has the potential to tap into the human computation resources available through the Web 2.0 model

4. Has potential to reduce USPTO's workload considerably if the knowledge of the system stabilizes to render the reviews reliable. Irrespective of that, wikipatents could be used as a tool by USPTO to reduce the time-to-grant, this system might just expedite the patent prosecution process

5. Tag based reviews of patents. If these tags are reliable,  the system could help the USPTO office to classify patents differently depending on how reviewers tag it. This could serve as a new way of classifying patents in addition to the traditional  classification methods

6. Market Research feature - A community based valuation of the patent from a market research perspective

Cons:

1. Reliability of reviews - It is very tough to model the reliability of this system as a whole. People could argue that, since this mimics the wikipedia model, it guarantees a certain degree of reliability. But the issue is that, there is a lot more at stake here in comparison to the wikipedia model. Hence reliability of reviews/reviewers becomes that much more important. If the lack of reliability scales, it could be a serious problem

2. Recognition among wikipatents reviewers - One of the main reasons wikipedia is popular is because it has a cult-like following of content contributors, who have carved their own niche and derived a lot of recognition and satisfaction in the wikipedia contributors community. There is a sense of accomplishment (and recognition) associated with each contribution. This seems to be lacking (or is not apparent) in the current wikipatents portal.

3. Knowledge of reviewers - All the reviewers are treated equally, in other words, all the reviewers seem to be getting equal weights in the valuation system.  There is no way (currently) to determine how knowledgeable a particular reviewer is. The authenticity of the reviewer and his mental acumen of the reviewer will play a large role in ascertaining the reliability of the system. Hence this will be a major challenge faced by wikipatents

4. Inventor participation - The inventor seems to have a very small role to play in the system. It seems very important that the inventor participates in promoting his patent or help provide an accurate translation of his idea in order to help reviewers do a better job. This seems lacking currently

5. Stability - This is closely related to the reliability issue. There is a certain lead time involved for this wikiproject that will render the system largely useless until there is enough confidence in the system, such that USPTO starts using the resources in a productive way. The rationale might be that something is better than nothing. If it is that, there might be other ways of achieving a better 'something.'

Conclusion:

Wikipatents is definitely a very impressive undertaking. The possibilities are immense and largely appears to be heading in the right direction. It could well serve the purpose of expediting/improving the patent prosecution process by the USPTO.

On the flip side, there is a possibility that the usefulness of the system might hit a wall beyond which it may not be any more usable. This could be attributed to the reliability issue and the concern over the quality of the reviewers.  They should definitely consider tackling these issues with utmost urgency. On the whole a very interesting concept, a kind of concept that needs proof to validate the vision.