CVA3660 (and maybe other Nespresso models) Fault 72 & Fault 75 root cause & fix

The subject of this 'fix' is problems with capsule feed (Fault 72) and capsule eject (Fault 75). After about two years of unreliable operation with these faults occurring fairly frequently, I think I have finally diagnosed the problem and with it a fairly crude solution which, fingers crossed, does seem to be effective.
The particular capsule feed fault I have experienced (doubtless there are others) was that the capsule would not move fully onto the capsule plate as it is 'hooked' from the carousel through to the plate. With the capsule thus slightly misaligned, the brew head moves down from above towards the top of the capsule with its usual strong force, but instead of enveloping and clamping the capsule prior to the pressurised water activation, its metal rim comes down on the top of the misplaced capsule and crushes it. Fortunately the machine detects this and aborts the cycle rather than attempting to pump the pressurised water as normal. This flags up in the display as Fault 72: the machine comes to a grinding halt and can only be reset by removing the carousel, capsule plate, full power off etc. And it can sometimes be very difficult to remove these parts as everything tends to be jammed together after the capsule crush.
The other persistent fault was failure to eject the spent capsule once the coffee is successfully dispensed. This flags up in the display as Fault 75. Again, the machine comes to a standstill and demands the removal of the capsule plate and full power off. Fortunately this fault is much easier to clear: wiggling the capsule plate as it is removed usually dislodges the capsule backwards into the chute which leads to the waste container.
Both these faults were occurring depressingly often resulting in many wasted capsules. It had got to the point where I had to remove and replace the capsule plate to ensure its swinging platform (the part of the capsule plate that swings up and backwards - Miele call this the 'hinged ramp') was seated properly before every attempt at a coffee, and even then I sometimes still had the faults. For many months I thought it was due to the swinging platform on the capsule plate not descending fully to its rest position after a capsule eject cycle due to stickiness or dirt on the plate (this is one 'fix' that is sometimes discussed online). The grooves on the swinging platform that the rim of the capsule rides in are narrow and even a slight misalignment would upset the smooth feed of the capsule - but even with brand-new capsule plates the fault mysteriously persisted. I also suspected that the white nylon hook which grabs the capsule and moves it onto the plate might be misaligned somehow.
The Miele Nespresso machines (CVA3660, and maybe other models) certainly seem to have a design weakness around the capsule feed and capsule eject functions - reading the (very difficult to find) Service Manual for these machines reveals several design modifications around this area. I have had two Miele service visits for these faults, both of which were very disappointing. The technicians were obsessed with the possibility that I had used cloned capsules (I hadn't) or that the capsules were bent or distorted in some way (they weren't). Having argued this out, the best they could do was sell me a new capsule plate (which slightly improved things, but not enough). They weren't willing to dismantle the machine to delve further into the problem (and to be fair, I couldn't see any benefit in doing so given that all the invisible 'internals' seemed to be working correctly - it was the visible parts that seemed to be at fault).
The breakthrough came when I started to think about how the two apparently unrelated faults might be linked. The capsule feed fault is more subtle than the capsule eject fault, so I first looked at the eject fault in the hope that the root cause might be more obvious. Eject is accomplished by a thin forward-facing lever, lying horizontally along the left hand side of the cavity in the front of the machine where the capsule plate lives. (It is all very dark inside this cavity and I found the best way to inspect it was to remove the carousel and shine a torch sideways from the carousel side into the cavity, through the aperture where the capsules are fed in). There is a thin groove on the left-hand side of the plate's swinging platform which engages with this lever as the plate is pushed into position. The lever has a pivot at its far end, and the internal mechanism rotates the lever clockwise about the pivot (looking at the lever face-on) by around 60 degrees to raise the plate's swinging platform with the capsule on it, and thereby 'flips' the capsule backwards down the chute. For the eject not to work, I deduced that this lever must be failing to rotate, and indeed there are quite alarming graunching noises accompanying the failure suggestive of some sort of jam.
On closer inspection, the area around the lever's pivot was heavily encrusted with burned-on oily coffee residue, sufficient to make me very suspicious that this was jamming the mechanism. With care, it is just possible to get a small screwdriver to this point with access from the front of the cavity and from the right hand capsule aperture (as per the torch comments above) and carefully scrape away around the pivot. It was very satisfying to see large flakes of fossilised coffee residue come away from around the pivot, just at the point where they looked like they could well be impeding rotation. I then vacuumed out the whole cavity with a small nozzle.
I was delighted to find that this seems to have cleared the eject fault, and even more delighted to realise that the capsule feed problem seemed to also have been solved. This is because the feed problem was indeed caused (as I had thought) by the swinging platform not seating back properly into its rest position - but the root cause of this was not a dirty capsule plate but rather the lever struggling to drive back to its rest position due to the jamming problem (even if it had successfully managed to eject the capsule), and therefore holding the swinging platform slightly above its rest position.
Clearly the machine is crying out for a full disassembly in order to clean the pivot properly, rather than the superficial poking about which is all that can be done otherwise. I may embark on this one day if the problem returns, but the Service Manual makes clear that stripping the machine down to the level needed for this access is a pretty fearsome task.