For the ECML, I'm talking about normal locos, with special couplers if necessary, not special locos. Assuming Lincoln is the only ICEC route remaining off-wire in 2025 you'd need just one loco diagram in England and a very small number added to the highland sleeper pool for the East Coast Scottish runs. Alternatively you could go the bi-mode route and share the pool of 13 (if I recall correctly) East Coast bi-modes with the MML TOC, who also are likely to have a few off-wire services. But, on it's own, ICEC shouldn't need a diesel fleet. As for that £200m, most of it is probably time savings over attaching/detaching the drag loco, which DaFT thought would take far longer than it actually would.
Your persistence in using the term 'DaFT' throughout this thread suggests you don't have the ability to competently contradict their thinking, and is becoming annoying.
Special locos just means locos that are designated to run these passenger services. It's not clear to me whether or not this could be achieved using something we already have, but given we're talking leasing it anyway, I'm not convinced it actually matters.
Whilst it is true that there could be the potential for cost savings by pooling with the sleepers, this is only one or possibly two out of maybe six or seven, so isn't as substantial as you might think, a lot of the cost being the baseline for having any in the first place.(Although it would greatly improve the apparent financial situation of the sleeper.)
Note the £200m must also include GW routes. The extra expense will be all the standard reasons why widespread use of locomotives and coaches is fading out, particularly the higher track wear over these (quite long) sections of route, operational complications from adding/removing locomotives and the need to maintain competencies on a wider range of things.
I didn't say IEP did I? I said bi-mode. Can't Alstom put diesel engines in a new Pendolino? Virgin chose loco-drags over trying to obtain a bi-mode tilting train or sticking with diesels under the wires.
Designing a bi-mode tilting train suffers from the same expense problem as designing a tilting version of IEP. In fact, they must be very nearly equivalent statements, so I view this response as petty, picking up on a very weak technicality, rather than in any way showing the argument to be false (just in case you were wondering

).
Bi-modifying (see what I did there?) a small fleet of something based on a Pendolino would be very expensive due to the amount of important and already quite tightly packed stuff that would need moving around, which is surely the underlying reason behind not considering it.
That's why I gave up suggesting loco-drags for the Cotswolds line and accepted bi-mode IEP for that route.
The 67s manage the Aberdeen and Inverness routes on the sleeper don't they? And EC's Hull, Harrogate and Lincoln services don't have all that many stops while away from the wires.
Working in reverse order here: These last three are the most sensitive EC routes to extra time for needing to couple on a loco and potentially even run it around at it's destination, so a bad support for your argument.
67s working fine on sleepers is irrelevant as the Sleepers remain as locomotives and coaching stock due to the routes they have to operate and the current lack of a financial basis for providing something different as opposed to just carrying on. EC didn't have these constraints when ordering new trains, and lo they went for bi-mode.
I'm confused as to why you now accept bi-mode for the Cotswolds but not for very similar routes (mainly less than or equal to 100mph, plenty of stops, etc) - you then lose quite a lot of the benefit by only having one standard thing, which surely underpins your whole argument for an all electric EC.
Very wrongly, I am utterly dismayed at their decision to scrap 91s 10 years early and if I can make them change their mind I will.
I'm intrigued to know what difference you think you can personally make.