A 4am start in Glasgow gets you on a plane at 6 something and on the ground in the South West an hour later. A 5am train will be somewhere near Wigan. The time saving is so large that trains cannot compete, at least not until the environmental impacts are properly costed.
A direct train Glasgow to Bristol would take 5:30, with the need to be at the station (being pessimistic) 10 minutes before departure.
After HS2, even going via Old Oak Common, there's potential for this to be reduced to circa 5 hours which could be about +2 hours on flying (which sounds crazy, but it could be the quickest way, assuming a reasonable connection time) and potentially puts Exeter about the 5:30 journey time vs the circa 4 hours for flying from Bristol (with a typical 90 minutes drive from Exeter to the airport at Bristol).
Where rail could aid with reducing overall journey times (a little) is where it could start/end closer to where someone is going, so not needing to go through the process of hiring a car (instead using a taxi for the last mile or so from the station).
Other factors may come into play; service frequency would be higher with rail, like if there's a delay rail parts pays out after a much shorter delay than flying.
With many more businesses being keen to be shown to show their green credentials and being carbon neutral, it could well be that rail does start to make inroads into flight usage.
However sleepers aren't likely to make much difference on that and any inroads are likely to be limited unless there's significant further reductions in journey times (HS rail further north than HS2, HS rail West from London and/or HS rail Southwest from Birmingham).