Sanj.varah, I've very grateful for your explanations. I think I follow most of it, but there's at least one matter that still puzzles me.
If a certain length (including allowance for fittings, etc.) of 22mm pipe has a pressure drop of say 0.5 mbar for a given gas flow and a much longer length of 28mm pipe has the same pressure drop of 0.5 mbar, then when the two pipes are joined end to end the total pressure drop for the given gas flow will be 1.0 mbar.
However you seem to be saying that it will make a difference if the 22mm pipe is upstream of the 28mm pipe rather than vice versa (due to "recovery"). I can't see this, since the resistance of each pipe is exactly the same. Is it due to greater turbulence arising at the junction between the two pipes where the smaller (higher velocity) pipe leads into the larger (lower velocity) pipe?
If the transition from higher velocity to lower velocity causes more resistance than vice versa, would it not depend very much on the smoothness of the transition and would it actually be significant for a typical in-line reducer? Also where the outlet from the gas meter is 22mm anyway would it then make any significant difference which pipe came first?