Any mutations that changes the disease virulence (newly learnt word that Grant Shapps doesn't know) are entirely random.
It could happen with any of the 50,000 current strains at any time.
Changes to how contagious it is, does matter. As thereafter, future mutations could potentially (entirely accidentally) cause worse virulence.
So you then have a very contagious disease that is hidden in hosts for 48 hours, and in some completely hidden, which causes more devastation in some people than the previous one.
So any hint in data that one is becoming especially dominant (esp in small social germ spreaders!) should be taken seriously until the full picture is known.