Comjuice is like Alexa, only far less accurate
I tested the service using my own company's domain, because I'm in charge of Analytics and I know exactly how many visitors we have, as well as just about any other statistic. What Comjuice gave back was so inaccurate, it could have been made up for all it's worth. It was a tiny fraction of the actual traffic, and included completely random statistics such as "Estimated ad revenue" (which had numbers in it), when we don't even carry paid ads (it's a site for a business company). The "Unique visitors" graph count was also completely wrong, although this comes from Compete (still, why include it if it's so inaccurate?).
Possibly the metrics which are easier to figure out were correct (number of pages indexed by Google), but even that was totally misrepresented; it said we have "3,400 pages" indexed by Google, when in fact 95% of these are not even real content pages and are clearly marked as such on the site (totally different system, and different subdomain). So Google may index them, but Comjuice sure isn't smart enough to tell the difference; had it been a competitor of mine (and not my own company), I would have gotten a completely skewed picture of what their site performance is like. This is one tool not to trust.