The stock that seems to form 5T74 23:13 Salisbury to Eastleigh which then forms 2T75 00:30 Eastleigh to Portsmouth Harbour as
@swt_passenger actually arrives at Salisbury as 2S67 21:07 Romsey to Salisbury at 22:15 so sits at Salisbury for a good hour.
Now I have four alternative uses which I believe are more economical ways of getting the unit concerned from Salisbury to Eastleigh which are:
1. 1L69 21:20 London Waterloo to Yeovil Junction runs with two portions dividing at Salisbury with one portion continuing onto Yeovil Junction and the other portion continuing onto Portsmouth Harbour via Eastleigh.
Now 1L69 is booked to arrive at Salisbury at 22:55 and depart at 23:03 so is enough time to effect the split of units here.
2. 2T75 00:30 Eastleigh to Portsmouth Harbour is rediagrammed to start back from Salisbury using nearly the same timings as 5T74 23:13 Salisbury to Eastleigh.
The paths are already there, it's just instead of running it ECS you are running it in passenger service which means more revenue opportunities.
3. Use the unit off the Lymington branch say 450084 which terminates at Brockenhurst as 1J66 22:36 Lymington Pier to Brockenhurst why not either extend that working to Southampton Central or run it ECS to Southampton Central and instead run a Southampton to Portsmouth Harbour via Eastleigh service.
This isn't perfect but means you are not operating a DMU over a purely 3rd rail route.
4. 1B77 23:05 London Waterloo to Poole runs with two portions dividing at Eastleigh with one portion continuing onto Poole and the other portion continuing onto Portsmouth Harbour.
Now 1B77 arrives at Eastleigh at 00:22 and departs at 00:23 so by retiming this service to have a longer dwell time at Eastleigh you could operate both 1B77 and 2T75 as EMUs.
All are economical ways of getting a unit in position with both a decrease of DMU over 3rd rail mileage and a decrease of ECS mileage too.