Forgive me if I have missed it, but I have never actually seen any official reference to 'safety' regarding the decision not to progress a 'pod' seating option. Do you have a link at all?
Putting that to one side, however, I can see that it's more than likely that economics and commercial reality would put-paid to the idea of 'pods' anyhow.
If a comparison is made with longhaul post war air travel, broadly-speaking, all passengers started out being accommodated in the same kind of seats. After a while, airlines started making some seats smaller and some a bit bigger. There was a hiatus around the advent of the wide-bodies, the L-1011s, DC-10s and 747s, with some short-lived dalliances with lounge type seating, but the 70s oil crisis, deregulation, competition et al meant that seating densities overall largely reverted to higher levels.
In the early 80s and onwards, airlines started introducing premium 'cabins' with much larger seats, offering significantly more recline.
In the 90s, the first commercial airline flat bed 'pod' seat was introduced in First Class, as aircraft were by then demarked by up to four cabin classes. During the early 2000s, flat bed pod seats were steadily rolled-out to become, largely, the Business Class norm.
As this happened, many airlines took the opportunity to retire their First Class products, as it became increasingly difficult to maintain a commercially-viable gap between two flat bed products.
Accordingly, longhaul airlines started at the opposite end of the scale from overnight sleeper trains in terms of accommodation density and, although the notion of multi-level couchette type bunks on 'planes is floated every now and again, they have thus far not been progressed as they're largely incompatible with the way that airline service works.
The railway in this conext, however, started not just with multi-level sleeping platforms but, almost from the outset, with private rooms. Multi-level sleeping being, generally-speaking, an extremely efficient use of floor space and one that can be priced accordingly.
So, even if there was a safety edict, imagine a pair of airline Business Class flat beds, 6'6" long and about 46" wide, plus the pod 'cocoons' surrounding them. Now work out whether you could get one of those pairs sideways against the outer wall, under the window of a sleeper room. The answer: I don't think that you can, and you'd lose the washbasin too. If the safety angle isn't in fact true, then 'herringbone' arrangements would be a bit more efficient, but not massively-so. In a single aisle UK rail situation, you'd probably stick with 1-2 across the aisle, but you'd need sufficient separation at the foot of the pairs to avoid the 'climbover'.
To move to a pod arrangement in seated accommodation would therefore be less efficient than normal seating and quite possibly, multilevel sleeping platforms. The only way that that could be reconciled would be by pricing it accordingly. However, it would be significantly less private, and potentially a good deal less quiet, comfortable and secure than a sleeper room.
So, how would you price it? The quality of service would be below that of a room, but the land-grab could be a bit higher. The quality of service would be much higher than that of a seat, but the land-grab would be significantly higher. You could cannibalise demand from rooms, or you could find that you had to underprice them to such an extent that the seats are no longer viable and the real low-cost option disappears.
And how would you police self-upgraders who had bought seats and help themselves to a vacant pod? Without airline staff ratios (and it even happens on 'planes with them) it'd be a challenge.
On that basis, the whole notion of pods seems to me to be a complete commercial non-starter in the context of a sleeper train.
The only way that I could see them being relevant on a train at all is as a premium option on services which do not convey sleeper accommodation - which is effectively what Trenitalia and some Asian operators do with their high-end options.
To think about couchettes instead, within the UK loading guage, again I think that the economics are stacked against them from the operator point of view. Floor-space wise, you'd at most save the width of one washbasin width per pair of current rooms. So, that means that you'd need about 5/6ths of the floorspace assuming no narrowing of the sleeping areas themselves. If the operator is working on 100% occupancy of rooms, and it's £200 a room, and they can use all the released space to increase capacity pro-rata, then they'd need 100% occupancy of each couchette at £42 per mattress to make the same revenue.
They would, however, have to balance this against the opportunity loss of not being able to sell 0.83 £200 rooms per two couchettes to someone or a pair who wanted privacy, as well as the incremental costs of marketing to and fulfilling service for potentially 20% more passengers.
The only way that this would work, I think, would be if you applied the no-frills airlines business model and see each ticketholder as an add-on sales opportunity. To do that, you'd need to significantly increase the opportunities for high-margin 'on-spend' - which would usually be for things such as food and drink. However, again, you then have to assess whether the folks that you're attracting with lower prices are actually going to display a propensity to make up all the revenue gap propping up the bar. And if they do, you need more staff.......
On that basis, if it's a reasonably accurate assessment, I can see exactly why CS has specified the stock in the way it has, and I would probably have done the same.