The two sites I referred to in my article is http://www.freeforex.net and http://www.hilux4x4.co.za . They target totally different groups and types of internet users. The Forex on is the one with lower traffic yet at some stage its Alexa ranking was the same as the Hilux one (both were at about 220,000). As you rightly pointed out this shows that more visitors with Alexa toolbars visited the Forex site than the Hilux one.

The difference between 6,000 and 45,000 visits per month is significant, yet to the (uninformed) outsider it seems as if they receive more or less the same traffic.

Altho the Alexa stats is not completely useless. It is like any statistical data. It is just a snapshot and a projection of what it is or could be.

With any survey or statistical information, you can never query the entire pool. You take a reasonable random sample. The extrapolate that into something that might make sense.

Users who visit your site and who have the Alexa toolbar are probably similar to the random sampling.

With any statistical data you have to compare it with something else. Perhaps another set of stats, or perhaps history trending, or what ever, to see if your sample is correct or not.

There are many reasons why your two websites do not exhibit the same
stats.
Alexa is off
Your other stats are off
Your niche draws more Alexa users than the other blog
It is most likely that Technical blogs will be visited by technical people who have a very high possibility of having the Alexa toolbar installed.

Having said all that, any set of stats can be manipulates, check out my post on SEO stats. http://www.integralwebsolutions.co.za/Blog/EntryId/195/Should-you-believe-your-SEO-stats.aspx

But I would compare stats with the likes of Google Analytic, Quantcast, etc