My immediate thought was that Julian is right, but some sums never go amiss
the swept volume per rev is 105880 cu.mm. If the loco is doing 8mph at 100% cutoff, the steam speed down the proposed pipe is 23 m/sec, which is quite low (usual figure for steam pipework systems is 25 to 40 m/sec). With steam at 6.5 bara, the dynamic pressure (density*velocity^2 / 2) is 897 Pa (0.13 psi). The pressure lost going round bends etc is the dynamic pressure multiplied by the loss coefficient, which is usually less than or equal to 1, so the proposed pipe is plenty big enough. I'll bet it has a much larger area than the hole in the regulator. In reality you don't get 100% cut off and full speed, and so this sum has a lot of conservatism, a much smaller pipe would suffice, but not much advantage