Phil,
I think the problem is far more likely to be the dampers than the springs. The springs do of course resist compression, so a weaker spring will be more likely to allow the suspension to 'bottom out' than a stronger spring, but, the damper also resists compression (or more accrurately, resists the speed of compression, i.e. the more rapidly the spring is being compressed, the more the damper resists the compression). So the weaker the damper (or pair of dampers) the more likely the suspension is to bottom out, i.e. the spring will compress faster and farther for a given 'impact' loading because the damper resists less. The stock rated front dampers are quite weak even when new, and age does them no favours.
IMO the 147 needs front dampers that are at least as stiff (in 'bump') as the Bilstein B6. When I acquired my 147 I thought the front dampers were worn out, so fitted some new TRW stock rate front dampers, which were barely an improvemrnt on the discarded originals. I then fitted B6 front dampers, which were a huge improvement (if still not quite as stiff as I'd prefer). Stock rated rear dampers are not nearly as under damped as the front ones are (I also fitted new TRWs to the rear end, because the originals were well and truly toast, and found them acceptable if not quite as stiff as I'd really like them to be, especially in 'rebound').
Note that springs don't get 'weaker' as such. They can lose 'free length' (and so the car loses ride height), but the inherent spring rate doesn't change. If a new spring requires say 100kg to compress it by X, then the same spring when older and 'sagged' will still require 100kg to compress it by X. The only difference will be the static ride height, but, since the ride height is lower the suspension will require less compression before the suspension contacts the bump stop and bottoms out.
Spring rate is only affected by the shape of the spring, which includes the wire length and diameter of the wire (more wire length = softer spring so more coils = softer spring, and thinner wire = softer spring). Rate isn't significantly affected by the metallurgy of the steel from which the spring is made, not the carbon content, not the alloying elements, not the state of heat treatment, nor the 'fatigue' of the metal (unless the wire is cracked or starting to crack). The difference between two springs of the same dimensions, but one made from a 'good' steel and the other from a 'bad' steel, is that the spring made from good steel can be compressed further and many more times before the spring starts to develop a 'sag' (or less commonly crack, break).
Regards,
John.