Which is the precise point.
In terms of a worthwhile means of measurement, which can only be from a large sample of people with varied demographics over a long time (e.g. a generation or more) there is not definitive evidence one way or another in terms of outcome in quality adjusted life years.
It is known that statin drugs reduce undesirable blood lipids in certain groups of patients. To that extent, they do as billed.
We also know that there is some correlation between poor lipid profiles and poor outcome from cardiovascular events.
What we don't know is whether poor lipid profiles are causal for poor CV outcomes and we also don't know whether statins, using a broad study base over a long period affect outcome.
However, it's a very long stretch to say that statins don't work at this point. The most pessimistic honest view is that they may work.
To that point, a small sample taken from what happens in one bit of the ambulance service is not useful in any way at all as corroborating evidence one way or another.
The only valid way to compare outcome is to test across a wide population for an extended period, which is what is happening. If we adopted the principle of don't use a new treatment because it might not work, we would be in the middle ages.