*by which I mean "wrong".
As always these composite indicators are determined by the choice of the input data. And as is often the case they have chosen things that are easy to measure... A great example (which explains, in part, why Cambridge does so well) is the presence of a University. And another is the academic ranking of its Computer Science degree. And another is the number of citations coming form the Computer Science department. Good news that only CompSci counts for innovation in IT.
There are some interesting network metrics, but it seems they are based on things like joint patents and company ownerships - neither of which reflects startup innovation (it takes years to be granted patents, and usually you get bought up after you have innovated things, not before).
It would also have been nice to know what time period the data reflected, not just when the report covered...