what on earth is the logic of allowing a free choice of column diameter (subject to keeping the same linear flow of solvent), but only permitting smaller injection volumes? If someone runs a method on a wider-diameter column, they can (and probably should) inject proportionally more. The acceptable injection volume should surely be expressed per cross-sectional area of the column.
As you wish:
"When column dimensions are changed, the flow rate may be adjusted as necessary using the following equation:", followed by the proper formula including column diameter and length. One of the few thing Ph.Eur. got right.
The sentence "Injection volume: may be decreased, provided detection and repeatability of the peak(s) to be determined are satisfactory; no increase permitted." only refers to injection volume changes when the column dimensions are NOT changed.
By the way: "free choice of column diameter"? I'd wish that was possible. Only +-25% allowed by Ph.Eur. Which means you're allowed to change from a 4 mm to a 3 mm column, but not from a 3 mm to a 4 mm column. And forget to scale anything to 2 mm columns...