hmmmm the cynic in me says it may be more down to the fact they now use dedicated rather than shared rather than actual capacity as they imply. Whilst they arent lying by saying "Investing in their network", I wouldn't be surprised if its more down to logistics.
Dedicated gives them a lot more flexibility when it comes to routing, and they could set up regional gateways....
BUT what really strikes me is now that they are running dedicated then any possible points of congestion due to bandwidth issues will be at the MSILs which are dotted around the UK at the Core Node locations.
In days of old the ellacoyas/proceras sat by the Juniper gateways which terminated the host links. Obviously the hostlinks are the possible points of congestion for shared WBC. If there were problems at the gateway, then the ellacoyas could spring into action.
So the question is... how effectively would an ellacoya in London manage traffic when the point of congestion could be at a MSIL in Glasgow without affecting traffic from other regions.
I could be wrong, but ideally the ellacoya would have to be near to the MSIL to be most effective.... and the old gateway ellacoyas could in actual fact be pretty useless when it comes to MSIL congestion?
I'd have to think about it a bit more and do a bit more digging, but off the top of my head I think they would have to steer traffic from certain MSILs to certain gateways, which doesnt appear to be part of their network design.