I've got to first say, Spamcop does indeed filter. However, their filtering is not exclusive to spam. You have to ask which is better, false positives, or false negatives. That SpamCops filtering is deliberately slanted toward false positives is no accident.
There are several ways of trying to deal with spam. Among them there are filters, there are target specific addresses, sender verification methods, and temporary expiring addresses. While I've nothing but praise for personal blacklists filtering, and whitelist filtering, the weakest method of effectively avoiding spam is through universal filters. Filters do have an advantage though in that is one of the most effective in discouraging the source of spam. The other methods basically just eliminate the spam, but does nothing to discourage or inconvenience the spammer. While such filtering as open-relay blocking is more or less indiscriminate, it is targeted toward ones worthy of the inconvenience. Open relay is without justification, and a major contributor to the spam problem.
SpamCop though is an IP blocker. That would be okay if the the spammer's IP was blocked. But, SpamCop has attempted to take the "discouragement" factor inherent in open relay blocking, and apply it to IP blocking. They block entire networks. What's worse, they do so without proper investigation, and without regard to legitimate users that are dependent on the network, or ISP. The large segment of users that are tied in with SpamCop's filters, often without the users knowledge, means individuals, newletters, mailing lists, and even the entire user base of some providers are unknowingly sending email that goes nowhere. They don't even recognize the problem until all the mail comes bouncing back, and the network involved starts checking into why so much email isn't working. Lockergnome, Langalist, Infopackets Gazette, and various large, and totally opt-in lists have been blocked. These lists are those someone subscribes to, receives an email notification and responds to it. They get information on unsubscribing to the list, which is usually as simple as clicking a link or sending an unsubscribe request by email. Many of those wishing to unsubscribe though apparently choose to attempt to unsubscribe by sending a report to SpamCop. Due to the irresponsible manner in which a blocked IP is determined by SpamCop, the list is blocked. SpamCop is quick to block, and slow to remove the block.
A very large segment of the antispam community has pleaded with, and tried to reason with Julian Haight to change the method, but to no avail.
The integrity issue, is that Julian has attempted to fight competition by blocking the IP of competitors. Spamex, Emailias, Sneakemail, SpamMotel and such compete in a method that SpamCop can't. They don't filter. With them a special email address is generated by the user for a specific purpose. Then any email sent to that address is forwarded to the users real address. Mail doesn't origniate at those servers, but is only passed on. For SpamCop to block them is totally without justification, and is only a business tactic to destroy them.
If you want to research the problem, here's a few addresses. The pages also contain many links. It's enough to keep you busy listening to the same complaints for hours.
http://jhoward.fastmail.fm/spamcop.html
http://www.politechbot.com/cgi-bin/politech.cgi?name=spamcop
http://www.azc.com/htmls/faqs/spamcop_blocking.html
http://www.spamex.com/news/2002.12.16.spamcop.cfm
http://www.website101.com/SpamFilter/spamcops.html
http://new.theage.com.au/articles/2002/12/19/1040174329829.html
http://www.politechbot.com/p-04129.html
http://www.spamresource.com/sc.html