Similarly the direct reading compass was probably sufficiently far forward.
However several known potential effects were ignored. When more than one engine was changed we were supposed to do a compass swing. When there was a mjor component change, ie the wing, or part of the system, or when there was a large change of 'magnetic latitude' eg a move from UK with a 40 degree dip to the equator with a zero dip.
Now the latter was ignored on transport aircraft and we successfully ignored it for bombers too. Similarly we made no correction for bomb loads either.
I remember once seeing a UK chart of local magnetic variation (1949 survey IIRC) that showed maximum variation around major cities like London, Birmingham and Manchester. This would suggest that there was a significany change in compass accuracy near the cities although this would reduce with altitude.
As far as bombing accuracies were concerned, a 0.1 deg error between true and magnetic (allowing for variation and deviation) would translate to a bombing error of only 51 feet for an offset aiming point 5 miles from the target.
In navigation terms this would have been insignificant as other unknowns - start position, fixing accuracy, wind, speed etc would all have had a greater effect.