If you can run two cables to the new point instead of one, you could incorporate it into the ring. The existing incoming ring cable would be extended through to the new point, then run back to the existing socket, where it would rejoin the existing ring, and pick up the existing spur too. That is the best configuration.
If you cannot run two cables to the new point, then we must re-open the debate about taking two unfused spurs from a single point, or rather from two very closely adjacent points. This has been discussed not long ago in a very contentious thread, where a specific configuration was proposed to bypass the intent of the regulations while claiming to comply with the letter. There have also been extensive threads attempting to discuss this point, especially on other forums, which have been derailed by misunderstandings and ignorance.
It is often pointed out that you can feed any number of points from a fused spur, i.e. where a single fused connection unit in the ring or spurred from it, feeds a radial with unlimited points with a maximum load of 13A total. To use this method in the OP's situation, both the new and the existing spurs would need to be fed from the FCU, not just the new one. However, although compliant, I consider this to be an inferior arrangement because the maximum load is even less than available from one (compliant) double point on an unfused spur, and there is a possibility of the 13A fuse in the FCU being subjected to long, low overload which is undesirable due to the tendency to cause heat damage.
Suppose there were two existing points in the ring next to each other; it would be technically compliant to take one spur from each. It might be poor design if one could foresee that both spurs would have heavy loads, as a significant load would then be concentrated in that area of the ring. However depending on the lengths of the cables and how they affect the load spreading, the loading of the original ring legs might be no worse than if the new points were incorporated into the ring, which would be considered best practice. Therefore, unless one is deliberately calculating cable lengths and resistances to avoid uneven loading, it is hard to argue against adding two unfused spurs each feeding one point.
I therefore suggest that it is not the end of the world in the OP's situation, to leave the existing spur connected to the socket outlet, and to take another spur from a junction / set of Wagos adjacent. I welcome constructive criticism of this approach. Please note that I am not advocating:
- Feeding two points from one unfused spur (non-compliant as cable is not adequately protected)
- Connecting two unfused spur cables into socket terminals through which the ring passes (likely to exceed terminal physical capacity, often 3x 2.5mm² max)
- Feeding known heavy load concentrations (e.g. socket-outlet cluster for dryer / washing machine / dishwasher) in this manner.