I never really took Alexa very seriously because of how easily alexa ratings are skewed depending on the type of website you have. It can be an indicator of what type of traffic you attract, but not necessarily an indicator that you attract a lot of traffic.
You may have a website that doesn't attract a user base that would typically have the Alexa toolbar installed. In theory, you could have a relatively decent traffic volume, but show an alexa rating around 3 million. The lower your Alexa rating, the better your score (kinda like golf). On one of my first eCommerce websites, I had a pretty decent Alexa rating which suggested to me that many of my shoppers had the alexa toolbar installed in their browsers. I found that quite interesting. With that said, my alexa rating on my AmazingCloaker site is a modest 466,617 as of the time of this writing. AmazingCloaker.com is about 11 months old. My handbag webstore was less than 3 months old when it's alexa rating was around 96,500. Hmmm.
Richard Lee wrote an interesting blog post which details his findings with several other alexa alternatives. The alexa alternative that I personally am most familiar with is Compete. As Richard points out, they claim to use an actual scientific formula. With a combination of ISP logs, Toolbars, ASPs and Opt-in panels, they claim much more accurate results as opposed to Alexa which only uses data from the Alexa tool bar.
Alexa by comparison would give me skewed traffic results on my eCommerce site vs. my AmazingCloaker site. If both averaged 1000 visitors a day, for example, and 600 of the eCommerce visitors had the Alexa toolbar, but only 200 visitors to my URL Cloaking website had the toolbar, then the Alexa rating would be vastly different even though the hits per day were the same.
Surprisingly enough, however, as your Alexa rating reaches the upper percentages (say 45,000 & below) your accuracy is also increased. Because of the shear numbers of visitors, the ratio of tool-bared visitors will equal out to more accurate results regarding superior traffic over say a site with a 100k+ Alexa rating. I believe that this "ratio factor" in high volume is what ultimately equals it out. Alexa & Compete are the two that I pay most attention to. I would agree that overall they offer the best indicators of how you are doing in the traffic scene. I also have the Google page rank indicator installed on my Firefox browser which I like to monitor as well.
I think that checking out alternatives in anything is normally a good thing, but for now I think that I will just stick with Alexa & Compete.