132.3 starts the reference as to splitting into circuits. then we move to 314 and 314.4 says no yet if you look at Appendix 15 it say an unfused spur may be connected to the origin of the circuit in the distribution board I would assume that is only for ring mains and not other circuits.
Define final circuit! If a 20A MCB feeds two fused switched spur units it could be considered from spur to immersion is final circuit not MCB to fused switched spur unit.
One can debate many of the regulations in the same way. So one has to look at does it cause a danger or inconvenience and make up your own mind.
Convention has it that you only run an immersion heater from the MCB and moving away from conventions can in its self cause danger where those you follow you jump to conclusions here you must consider 514.1.1 and 613 (d) (xii) and ensure good labels.
Sorry not a simple yes or no if it were me I would only connect two circuits to a single fuse in an emergency until work required could be done correctly.
this question arises when wishing to feed a central heating boiler system from an immersion heater supply as it is very unlikely to be supplying both together.
The regs book gives 3 'standard domestic circuits' to save us having to work out the loadings etc. ie Ring[A1], 32A 4mm2 radial [A2], & 20A 2.5mm2 [A3] radial.
We are totaly at liberty to design other non-standard circuits as long as if questioned we can show design parameters worked to, this of course is normal non-domestic practice.
The point you are looking at is if say a 16A mcb is used to supply even 4 radials from its origin at the board could you overload any part of any wiring & the answer has to be no! The total load you could impose on all or any of the cables would be 16A before the CB tripped. This is IMHO not a problem if no other choice easily available, as long as you feel happy to 'defend your corner' if questioned, I have lost count of the times I have seen 2 lighting cables put together on one 6A cb to free off a space in Con. unit for a shower etc.