Author states that Big O notation is to measure algorithmic complexity, that's simply not right. It is to measure runtime - a completely different metric.
The article states that Big O notation is used to analyse algorithm complexity and that's correct - it's to analyse the time complexity of an algorithm.
Even then, it is better to say that it is used to compare multiple algorithms on complexity.
Also, one of the things I found actually reading TAoCP is that in pretty much all algorithms, he gives a full instruction count of everything. So, it isn't like these preclude each other.
