Let me add some details here.
It may be important to download and look at a copy of the relevant patent, as Fig. 2 in particular shows you very much how the thing is laid out.
Yes, the patent drawing shows the worm gear from the Cardan-shaft drive at the bottom, while 611 has it at the top.
The view of the cambox drive end is not 'missing' a piece of mechanism; refer to the figure. The 'cruciform' piece with the nut in its center is the piece marked 66 in the figure, it's the 'driving' end of the cam stack. The four cruciform 'dogs' engage in the four visible milled slots in the end of the driving sleeve, which is shown simplified as 76, with its driving gear 79. I believe this piece has been fabricated from two halves, by welding, and this accounts for the two 'marks' 180 degrees apart that are visible on the end of the sleeve. (I also believe that the large driven gear is fabricated in one piece and then attached to the sleeve as required... you'll find that out as the cambox end is disassembled...)
I note that the cam stack in the picture is considerably 'outboard' from where it is illustrated in the figure, which to me indicates that the stack was physically shifted fully against the 'outboard' internal stop when the picture was taken.
Regarding what is 'missing' (other than the three large slot-head bolts): the patent drawing illustrates sleeve bearings for both sections of the driving sleeve itself, with the rear bearing possessing a lip that takes inward thrust on the driving gear, and a shorter forward bearing that also takes outward thrust on the gear. The clearances and fits of these sleeve bearings appear to be set by the assembly of the driving-gear case itself (the much larger bolts) and therefore there may be no more structure 'missing' with regard to operation of the system than there is for missing dust caps over the outer bolts in freight-car roller-bearing journals. (In fact, the patent drawing appears to show a slight clearance between the sleeve and sleeve bearing and the inside face of the 'cap'.)
On the other hand, there are an awful lot of screws in that plate for it just to be a partial cover over the end of the cam stack. And I, personally, would prefer to keep the driving dogs and channels both clean and well-lubricated, neither of which is remotely assured if the locomotive is worked for any length of time with that end open... so I would check the available sources to see if there was a plate or diaphragm there.
I do not think precise alignment of the U-joints in the shafts is particularly relevant. I've always thought that the use of the split shaft, with the very long 'center shaft' running in rigidly-aligned bearings, is to allow reduction of the unsprung mass at the back end, and reduce torsional whip or other undesired moments.
Remember that the gears themselves determine the 'physical' alignment of the cam with the drivers, once set. The shaft can be bolted up in ANY position, once the cam on that side is properly aligned and the locomotive is rolled first to a good dead center (via a precise method like the 'tram') and then advanced the proper number of d/m/s to correspond to observed valve opening. In practice, there will need to be a little 'compromise' because you must bolt up to fixed bolt positions in the flanges, but the amount of mistiming introduced by a 'fractional hole' even in the wrong direction is not likely to be critical. Likewise, while there has to be a splined joint 'somewhere' in the rear driveshaft, its assembly does not pose any particular difficulty (as, again, you have your choice of adjacent bolt holes when joining up to the flange on the back end of the center shaft). I'm sure there's some 'best' order for how you align the different sections for best performance... but getting all the universals exactly in line is, I think, immaterial too -- there's a certain air of military precision to universals all in a row, just as it's nice to see that all the hardware is security-wired as on a good gas-turbine engine, but is that level of sophistication really "necessary" on a 2-8-0?
I have not yet seen any discussion of the air cylinders that shift the mechanism. The patent drawing illustrates the 'shifter fork' as being centered between spring plungers, but the claims section specifically mentions, as a separate claim, that a double-acting piston can be used instead. I am waiting with some interest to see what is actually there, and how it 'exhausts'.
It would have been 'technically' possible to extend the cambox completely across the locomotive and drive it only from one side (the 'long-compression' Franklin patent illustrates this cam layout). In the present setup, the camshafts are much shorter and hence less subject to deflect or whip, but you have a separate 'drive transmission' on either side with no shaft across the middle -- my own prejudice would be to have at least the possibility of dogging a shaft in there for emergencies, after the manner of the V22 Osprey's 'redundant' shafting, but there is also the exhaust-tracting up to the nozzle to contend with; the current design easily fits in place of Walschaerts and piston-valve cylinders, even with the much larger admission tracting common to the Franklin System designs...
Implication of the two shafts is that you have two separate settings of running cams (with a borescope or whatever to line up initial valve opening) AND you have to roll the engine 'exactly 90 degrees' to bolt up the shaft to the second cambox after the first side... the right side, if it's leading... is done. A little more tinkering with the tram and opening up the tracting... but not that hard.
Meanwhile: the patent indicates, seemingly for more than mere explication, that all the cams have their 'keyways' machined for precision alignment on the shaft with just one position. I would not expect it to be difficult, at all, to matchmark the outer end of the driving sleeve with the housing, so that driveline fitting would involve nothing more than turning the universal at the cambox (to align the marks), setting appropriate DC (with the tram) plus whatever angle of advance is desired (again in d/m/s as desired - I suggested on another forum that a large degree wheel could be made up for this purpose, as the advance is a relative measurement) and then bolt up from the rear to the front, turning to 'make' the bolted connections in the least troublesome direction...
The gearboxes at the cranks are not positional encoders; you literally don't care where they are set (unless you're being officious and getting all the little universals to line up perfectly). Likewise, there is nothing in the driveline orientation that is required for timing. If you wanted to be persnickety precise, you could weld up the driveshafts so the flange boltholes were PERFECTLY aligned -- or use some other approach like microsplining the mating flange faces radially, and elongating one or both sets of bolt holes -- but I really don't think that's going to be critical.
It does follow, of course, that once the engine is together and under test, 'advancing' the cams in small increments is done very easily by slipping both shafts one bolt in the desired direction and retightening... repeat as needed, and then witness-mark the "best" positions everywhere.
It really does pay to look over 2,518,024 before speculating on what some parts of the design are supposed to be doing. In particular, the desirability and implementation of automagical *compression* control is as interesting as the more-obvious admission "cutoff" control (spoiler alert - it isn't done with fancy servos and feedback logic...)
RME
