The train preceding 6J37 is booked to leave Chester at 1822 (it's 1V98, the Holyhead - Cardiff, I know that only too well as it delays 6J37, which I drive on a very regular basis, at least 2-3 times a week, on average, by being late reaching Gobowen). The next train from Chester to Shrewsbury (1J70) is booked off Chester at 1917, by which time 6J37 should, if not delayed by 1V98, be setting back into the Kronospan yard (despite 1V98's frequent tardiness, it is still very rare indeed that the groundframe hasn't been handed back by 1930, with 1J70 not due off Wrexham until 1932, with maybe ¾ mile before CN940 signal). There have been a couple of occasions in recent months (both on Fridays, I seem to recall) where a late-running Network Rail test train was given the road south in between 1V98 and 6J37, leaving the latter waiting at CN940 signal until said test trains had reached Gobowen, but it is almost exclusively 1V98 which causes the issues. It's frequently late departing Wrexham even with a right-time arrival there, and I've no idea why this is the case - it somehow seems to happen even some nights the loco-hauled service is replaced by a unit, although perhaps to a lesser degree on such occasions.
TfW are very good at blaming "a late running freight train" for issues down there at that time of day, less so at explaining WHY that freight train was late. It is very, VERY rarely the fault of 6J37 itself - if for any reason it is likely to delay a passenger train of its own accord it is almost always held at Chester, either in the station or occasionally on the slows at Roodee. If it is put out 'in front' of a passenger train this is because it has been delayed by something else, which is, as I've already mentioned, almost always 1V98.
As for the bit of your post I've put in bold, 6J37 cannot be put/held in the loop at Croes Newydd, because the train is longer than the 1339-foot loop, so once it's past Roodee, it's not going away till Chirk no matter how long the previous train may delay it for.