The actual timings of 2F97, 1V45 and 1V46 were as follows:
Code:
. -------2F97------ ----1V45--- -------2F97------ ----1V46---
CHELTENHM 09:36 (1E) 09:41 (16L) 09:52 (RT)
GLOSBARNJ 09/42 (RT) 09/46 (16L) 09/57 (RT)
HORTONRDJ 09/44 (RT)
GLOUCESTR 09:45 09:49 (1L)
HORTONRDJ 09/51 (1L) <
GLOS YDJN 09/51 (1L) 09/47 (16L) 09/51 (1L) 09/58 (RT)
STANDISHJ > 09/52 (17L) 09/58 (2L) 10/04 (2L)
CAM&DURSY 10:03 10:08 (6L)
CHARFIELD 10/01 (19L) 10/15 (6L) 10/20 (7L)
YATE 10/06 (20L) 10:21 10:24 (8L) 10/27 (7L)
WESTRLGHJ 10/09 (21L) 10/27 (8L) 10/31 (8L)
...showing that 1V45 overtook 2F97 whilst it was calling at Gloucester, not really delaying it. The subsequent delay (4' between Standish Jn and Charfield) to 2F97 was indeed dobbed to following late 1V45, which is clearly just lazy and incorrect attribution (which hasn't been disputed). The loops at Haresfield and Charfield are both goods loops, and shouldn't therefore ever be used to put a passenger away for regulating purposes (the only provision in the regs is for when a failure is preventing the main being used, and then only to run a passenger straight through and out the other end). 2F97 was fairly tight in front of 1V46 leaving Gloucester, but wouldn't really have knocked it if it hadn't been for the extended dwell time at the two station stops, so I'd certainly suggest that the regulation was correct all round.