So does this mean Ashford would be the better intermediate station to bring back?
On costs it does. Of course the pre-covid growth numbers would suggest it's still not a good idea, but then there's the argument about the fact it was a pretty sparse service for all that time.
I originally would've said Stratford because it's like a London parkway station with an urban catchment
Stratford doesn't make any operational sense, and is the station with by far the least penalty to the potential market for not serving it. It's also not the London Parkway, that's Ebbsfleet.
All 3 stations have the problem that they're only sized for a fraction of the capacity of each train, so you need to start the trains from St Pancras anyway and therefore intermediate joiners are costing you potential fare money (no churn because of border rules).
Stratford: The market is actually limited largely to East/NE London due to other areas having better connections to St P., is well connected to St Pancras with relatively quick journey times by SE. It's way too early a stop on the LDHS services - it'd be like stopping all Avanti trains at Wembley Central or all LNERs at Alexandra Palace. It adds way too much journey time for what it offers, especially if the annual charges are c.£7.5m.
Ebbsfleet: Good for North Kent/SE london, well connected by road for M25 parkway role. however very little immediate demand, connections to surrounding rail services are pretty patchy (Platforms 5/6 are a good hike away from the international terminal), and it's still quite close to London so the option to travel in and double back is still present.
Ashford: Good for East Kent/Sussex, decent road connections for M20 parkway but limited scope for that role given additional driving time. Local Rail connections are good and well integrated to the station, it's also further from London to make the option of the double-back a significant negative factor. It's also closer to france so the journeys are more attractive for a day-tripping market if the right stop combos can be offered.
IMO, it's about even between Ebbsfleet and Ashford, but as Ashford has 1/3rd the costs of Ebbsfleet (largely, I suspect, due to the station being shared with NR), it'd be the sensible one to focus on. The question is whether you can stop enough trains there in a sensible pattern to have efficient staffing costs and still build a market that will cover your £3m/yr payments to LSPH and your own additional costs, and then whether doing so is more profitable than simply filling the train at St P. and running non-stop.
I expect the final point to be the achilles heel of any proposal to operate Kent stops, at least while the current border regime lasts (which shows no sign of relaxing any time soon).
but your numbers have got me thinking it might be otherwise.
On a pedantic note, they're not my numbers - they're the numbers the Station Owner provides as part of it's documentation.