I think you're a bit confused about what LINX is.
LINX is a neutral peering point which ISPs, etc, connect to in order to trade traffic for free. ntl and Telewest were both connected to this for some time.
This does
not mean that ntl and Telewest or any other member of LINX donate any bandwidth to other networks, it means that networks can be connected directly across the LINX switches without using an intermediate or transit provider. The connections on LINX are point to point and exist purely for ISPs to trade their own traffic. ISPs peer
at LINX, apart from a peering with LINX's collectors for stats.
What's been done there is that OVH have set up a BGP peering session with Virgin Media, both were already at and connected to LINX but the configuration wasn't in place, and are awaiting Virgin doing the configuration on their side, then instead of traffic having to go through Virgin's saturated DTAG link to reach OVH it will just go across Virgin's connection to the LINX LAN and reach OVH directly.
This should resolve issues people are seeing with speeds, so long as Virgin's LINX connections aren't saturated of course
Swift Diagram:
Before:
VM ---- DTAG routers and whatever else--- OVH
After
VM ---(LINX switches)--- OVH
Saves both VM and OVH a few quid / euros as well, not paying DTAG.