At the moment GPs are required only to confirm that the applicant does not have relevant medical conditions as defined by the Home Office guidance, or otherwise, which in turn the applicant has to declare on their application. They do not need to see the patient at all, just look at their records. If there is a relevant medical condition, the GP should say what it was and when it was, they do not have to give any detail about the impact on firearms ownership, that is for the police to decide.
As a professor of psychiatry, on publicly available information at the moment, this looks like a fairly typical family annihilation/autogenic massacre scenario, in the context of serious personality dysfunction, which would have been a longstanding issue, culminating in sudden complete decompensation and loss of control. It may not be possible to identify the final trigger(s) but it’s usually some sort of personal reversal. It’s rather disturbing to me that someone considered an anger management intervention would manage the risk in an allegedly assaultative young man who had a firearm like this one, but then again, FEOs do not have mandatory training in mental health, and mental health professionals rarely know anything about firearms. I occasionally teach to both audiences, but have encountered resistance from the NHS: the Royal College of Psychiatrists and the General Medical Council have zero to say about firearms, despite that ownership is not uncommon. To get back to Plymouth, Pete1774 is absolutely right that risks fluctuate, mercifully to this extent vanishing rarely. The risk of suicide by firearm is orders of magnitude greater than homicide, in fact homicide with a legally held firearm is so rare that the Home Office do not collect the figures.