I am no fan of the concept of taking things to the ultimate and have serious doubts about the interpretation of some parts of the (non statutory) regs I see here. I do however believe in the concept of thinking designs through, you are the designer and should be able to show mitigation for your design. In other words to use the argument that someone would have to not comply with the (non statutory) regs to justify your design doesn't really work. I would suggest you need to show, that the connected load could not be increased in other ways.
It's obviously impossible to "show that the connected load could not be increased", since it obviously
could be increased by someone who had such a desire, just as a 6A MCB could be 'upgraded' to a 32A one!
My main argument would be that the load could only be increased by changing a hard-wired fixed load, and that if one can't assume that a person doing that would not ensure that the circuit supplying it was adequate for that change, I don't know
what one could assume. I would say this was very analgous to showers - should we never be allowed to install a (currently perfectly adequate) 6mm² shower circuit, because someone might subsequently install a higher powered shower and 'upgrade' the MCB without also upgrading the wiring?
Out of deference to the views of those here (which, needless to say, come as no great surprise!), I'll probably 'give in' and feed the 3A FCUs from a 13A one - it only menas one extra FCU and a few extra inches of cable. However, I am in some senses 'annoyed' by having to do this, in relation to what seems to me to be a reasonable design in engineering terms, and which only goes wrong if someone subsequently decided to install different hard-wired loads without first checking that the circuit was suitable! Ah, well!
Kind Regards, John