The rule prevents negative outcomes for risky non-qualified patients from harming the trial, but the converse is also true. Positive outcomes for patients outside the regulated trial will not help gain approval at Phase II or Phase III.
On a board like hn, there's a tendency to look at all data as useful data, but this isn't true in the case of a clinical trial.
On further thinking, I'm not even sure what happens in the case of a patient receiving a pre-approval drug outside of the trial. Handling their data would likely need to meet the same standards required in trial (consent tracking, electronic signatures, etc) without any of the machinery of the trial to help. It could well be the case that nothing is collected if the cost to keep it and use it are too high.