Is there any evidence of a sword being made slightly wider than ideal with the idea that in use and with sharpening it would have edge material removed? I know some swords are difficult to categorize into an Oakshott number because of repeated sharpening and maintenence. Some seem to have wondered from XVIII into XV for instance. Would this be something taken into account by a smith?

Not to the point of being deliberately non-ideal straight out of the forge, no. For a tool you count on for your very life, performance right now is simply an infinitely more important consideration than performance at some theoretical point in the future maybe.