I think that you infer rather a lot about the design and build process that none of us can actually know. Whether the design was rushed or subject to lack of planning or foresight is something that we shall never know.
However, these trains are not brand new standalone products that can be subject to entirely new standards. They have to be made to work within the current standards for things like ETS and braking, and as such there is only a certain amount of leeway the designer and customer can expect before their product starts to fall outside of the standards.
The issue with the auto-coupler is not so much the number of connections but rather the need for a fully automatic system to carry the ETS as well as any control or communication circuits. The electrical connector block that you see on the Delner couplers of MUs is not rated to carry that sort of voltage (RAIB Para 98). As such, there was always going to be a need for someone to go between the train and loco and plug these connections in. I would also question whether or not it is actually any safer to have raised the position of the jumper connectors. I know that the Southern had a particular safety concern that caused them to have their jumper connectors at windscreen height, but I would question if it's safer to have your shunter on the ground plugging stuff in or balanced on a stepboard.
It is certainly possible for locos to use fully automatic couplers. ROG have some Cl37s with various drophead auto-couplers that they use when moving EMUs around. It's not impossible that the Cally Sleeper could have taken a similar route. Clearly you would need to have some sort of electrically controlled pneumatic brake with a main air supply and an electrical brake continuity circuit, as staying with the two pipe system would still require operation of the BPICs at either end of the train in order to vent the brake pipe to secure the train and to carry out the brake continuity test.
The downside of such an arrangement is that suddenly this train becomes non-standard, which is the sort of thing that normally provokes howls of derision from the members of this forum who see interoperability as some sort of Holy Grail. At present the Cally Sleeper can be hauled by ANY loco by means of a coupler adapter, but using auto-couplers and an electrically controlled pneumatic brake means that you either have to have a compatible loco or a translator vehicle to go on each end. Given the issues discovered with the 61 way connector, it's probably a good thing they didn't take this route, as controls over how and when you go about plugging in a jumper is a good way of ensuring that two locos are not connected at the same time.
You're right that I'm inferring a lot, and I might be wrong. I don't design trains but I do design other stuff (buildings mainly) and this means that I feel I have some sense of recognising things that tend to happen as a result of a compromised design process, though.
This really is just speculation but I can imagine how this project had a starting point of: we have these class 92s available, which can already provide a high level of ETS power, and this lets us have a new set of carriages, with high power demand, without having to build dedicated new locos. And I can see how that seemed to make sense, and allowed a bid where a sleeper with all sorts of power hungry stuff like showers and so on could be offered. And it could have dellner couplers to simplify the currently complex and labour intensive train splitting/joining operations.
But then I can imagine it progressing into detailed designs and all sorts of problems coming up, maybe to do with trying to fit complex plumbing systems into the restrictive UK loading gauge, maybe realising that this high power demand led to issues with how the automatic coupling could work, maybe realising that the elderly electronics of the class 92s worked in theory but not in practice with newer systems with different demands. A series of issues that all stem from the basic concept - high power demand newbuild rolling stock and refurbished locos. And I know how these kinds of issues can have cascading effects - you implement a workaround to one problem, which causes a new problem. And so on.
And many projects have a point at which they can go two ways: either, someone decides that there are so many unforeseen problems building up as a consequence of the basic original concept, the sensible response is to go back and rethink that original concept. Or, no-one makes that call, and the project crashes on, becoming gradually more complex and expensive and ending up with a result that is full of compromises and workarounds. To me, this has various hallmarks of the latter taking place (evident design compromises, lots of stuff that doesn't work well, loads of teething problems, extensive delays), and I can also see exactly what pressures could have lead to this (political pressure not to go back on the original promise, and so on).
I might be completely wrong of course.
I do wonder, though, at what stage was the design and position of the BPIC decided? Could it have been at a stage in the design where there was still an intention for a fully automatic coupling? Could that explain why the risk of it getting knocked during the plugging-in of jumpers seems to have been overlooked in the design process? Or did it get moved from its original position, in order to accommodate later changes? Could that have anything to with it being the "wrong way up" compared to other cocks?
By the way you say the CS stock can be hauled by anything with a coupling adapter at the moment - is that true? I thought it always has to have a 92 or 73/9 there, even if it isn't providing traction power. Effectively the 92 or 73 is the coupling adapter. Or maybe you mean for empty stock moves and so on? It appears to me that effectively they have ended up with something with no interoperability. So there would not be much lost if a dedicated coupling system were used.