This is not true.
If correctly designed both situations (RFC or radial) will have sufficiently low fault impedance Zs such that the breaker hits the "instant" magnetic trip point and then it will disconnect in typically 10ms or less. The disconnection speed once in the magnetic trip fault zone (rather than thermal overload trip) is not very sensitive to the degree of over-current.
In all cases for a typical UK home on TN supply the design has to disconnect in under 0.4s under hard fault conditions and in the case of MCB (and related RCBO) that always means hitting the magnetic trip point. There is some difference if fault let-through energy between MCB ratings (see below) but surprisingly small, but again the basic principle of electrical design is the cable size has been selected to meet both the operational current and the worst-case fault condition's heating effects.
As
@davesparks has already said the key is to have it
correctly designed, installed and tested so it is known that everything is meeting the regulations and nothing has a loose / high-resistance connection or dodgy insulation. Oddly enough this is an aspect where the RFC is better as the figure-of-eight test is very sensitive to even small excess resistance since every socket should be identical during the test so spotting a bad joint or socket with a dodgy switch, etc, is easier.
If you want multiple circuits to allow rooms to be isolated individually, etc, then radial make more sense than the RFC which is better suited to a whole floor, etc, that naturally forms some sort of a loop, but to characterise the radial as safer is completely wrong.
View attachment 104770