Had he run the Blitz test for 10 mins or more you would see the spike in traffic up, beyond what you think a Micro should sustain, and then it plummet to near 0 for a disturbingly long period of time[1].
If you are unlucky enough to have a Micro on a host that is fairly saturated, the performance you get is untenable.
Micro's are not "smaller than" Smalls -- they are a different type of monetization production from AWS allowing you to pay cheaply for little bursts of underutilized hardware.
There is no way (none... not possible, zero) that a Micro would be able to provide the bandwidth and CPU power to host a realistic site doing 10 million hits a day even if everything was straight from RAM.
Read through the EC2 forums for any length of time and you'll frequently find people coming in with reports of their machines "stopping" or "crashing, and totally inaccessible" for minutes at a time with a 100% ST ratio; every time it is a Micro that has been hammer for a bit either through benchmarking or use before the hypervisor puts it in a full nelson and brings it down to the point that SSH connections to the host cannot even be maintained.
-- I would also point out that not only is the CPU time allocated in bursts for a Micro, but the bandwidth is prioritized behind every other instance type (unless you were using a CDN, this would make hosting a typical wordpress site sluggish at best from a Micro -- and again, doing 10 mil hits a day... not going to happen in reality) -- yes you can offload all your graphical assets to a CDN, but now this article is about a $15/mo server and a $4217/mo CDN bill which is a very different article.
Additionally, if you need to use EBS at all in here, the story gets even worse with Micro's (even using something like RDS which requires network I/O in addition to hosting site content is all going to collapse on itself within the first 5 mins of the site's life with traffic like that).
All that said, the tips in the article are great. I only mean to clarify expectations with the use of Micro's. A whole swarm of them grinding through a work queue in random order is great; using them as the backbone of your web presence will have pain points. (Yes you can put 20-50 of them behind an ELB, but at that point why not run a handful of Mediums or a few Larges).
Anyone with a Blitz.io account, please feel free to setup this exact configuration and run the benchmark for 1hr with the same load to verify the meltdown.
[UPDATE] Ewan, I am not knocking your article, I wouldn't expect most people to be familiar with the ins and outs of every cloud provider. The tips and techniques are great regardless of the actual performance on a Micro (and applies to anyone trying to scale WordPress). Just wanted to clarify for anyone getting excited that they can now run their Fortune 500 website on a single AWS Micro that there are nasty surprises lurking just under the surface.
[1] http://docs.amazonwebservices.com/AWSEC2/latest/UserGuide/co...