There is a small piece of copper leaving the meter box and straight into the house

My understanding was we don't need to bond this where the incoming gas supply was plastic

The criterion is whether the pipe entering the premises is an extraneous-conductive-part  or not.

With a plastic supply pipe it is very unlikely to be an extraneous-conductive-part.

The criterion is whether the pipe entering

Although logically perhaps we should read "entering" as "entering or exiting" as it's the conductivity of the pipe rather than direction of gas flow that's the issue. I agree in a modern single building installation, where the supply pipe is the only place where the gas system comes into contact with the general mass of the earth, a plastic incomer is likely to mean the pipework isn't an extraneous-conductive part. Even if it goes down into a concrete slab of a modern building, it's still likely to be above the damp proof membrane, so still effectively insulated from Earth. Older buildings where the (often lead) gas pipes get embedded in damp masonry walls and into solid floors without a damp proof, it could be another matter. Strictly speaking the service pipe might still not be an extraneous-conductive-part, but it might be elsewhere in the gas system where it does enter/emerge from earthy surroundings. Ideally you'd bond at each such position, but historically a bond at the intake position was sort of presumed to cover all eventualities (interior gas pipework have to be all metallic anyway for fire reasons, so unless there's a deliberate insulating joint, it's was fair enough assumption). Just something to keep in mind if you're omitting the main gas bond.

   - Andy.

Well - the wording of the regulation for bonding connection location is "as near as practicable to the point of entry of that part into the premises" and of course applies only to extraneous-conductive-parts so I'll stand by what I wrote.

regs wise, correct.

regs wise, correct.