What is the level of danger you think this poses ? - that will depend what else is on that RCD, and perhaps on the power level and installation method of the PV inverter.

Are other things wrong?  - Are the  isolators and so forth suitable, is the overload protection appropriate etc.

Are there also batteries as part of the installation ?

Mike

Provided the RCCB switches N as well as lines (551.7.1 (ii)) - what's the non-compliance?

Depending on the inverter, a B-type might be needed, but that's a separate issue.

   - Andy.

Probably need more information/context and also maybe some pictures.

When was the CU/DB installed?

When was the solar PV installed?

Was it installed by an accredited MCS approved installer?

Is there battery storage?

Other points of note

Is there Type 2 SPD present?

Is there AC and DC isolation on the PV install?

Type of cable used for array and is it clearly labelled up as DC power

Is there a copy of G98?

Personally I think that Solar PV should be fitted to a fully RCBO populated board (CU/DB)  with SPD.  Caveat bi-directional.  This is as a starting point for Diversity and Earth Leakage.

Are you thinking along the lines of DC currents present and masking fault currents, preventing the RCD from operating?  

If the inverter is capable of DC feedback 

Surely the PV output (± battery) does not have to join at the origin?

If the PV array shares a distribution circuit, does it need RCD protection at all?

Does it come down to the inverter having simple separation. Transformer-less inverters may be able to feed a DC fault current into the AC side (Reference COP) The DC element May blind the split board RCD and potential leave other circuits without RCD protection.