With the brass rack, it may help to just play with it for a while to understand how the "centering" wires guide the whole shebang up and down in a parallel motion, smoothly and easily in their respective (brass?) holders, which are usually riveted to the front steel plate (I think it's the bottom one) and screwed in (?) to the plate on the top. This is so you can unscrew to get #41 out for replacement or to check (twirl it in your fingers) that it is dead straight between the centering wires, fix it, etc. As explained, the wire part on the front with the funky bends is what "drops" the rack (the tip of it hits right in the middle step of the snail surface) to set up the rack for the right number of strikes after actuation.
For the numbered illustrations, you're welcome. At least we can use them to talk. Another reference you may find useful is this:
Clockmakers Aid
For me, the best way to guesstimate pendulum length is to do a train count, plug in the numbers and then test it before cutting anything. It is not possible (for me) to know precisely where along the spring the "theoretical top point" of spring flexion or pivot is, so it's usually longer than the number that comes out of the equation. So, it's really a guide to better trial and error. The length also refers to the "center of mass" of the pendulum bob, so usually the bullseye, but on old pendulums the center of mass is not always the center of the circle. You just never know. You'll also need room for adjustment faster and slower, so go slow and test a lot before you cut.
When an arbor is removed incorrectly, the doofuss (I've done it!) usually bends the pivot. These clocks are built like a tank, and hardened, but not like the Parisiennes, so repair is usually not too bad, unless it's your first one. Go slow. If it were my clock I would look at all the arbors, pivots etc. and be sure none are bent, worn, scoured, contaminated and so on. Running it bent is not a good thing, even though they are very tough. If it is bent, try to align the bend in the direction the plate will swing out --or remove both sides at the same time. A finger twirl or roll on the bench will usually be enough to see any bend. On watches, a graph of rate error can sometimes show a bent pivot. It looks like a sine wave as it speeds up and slows down. I would not regulate until all other problems are resolved. Like a bend, and (for me) a disassembly, inspection, cleaning, adjustment, oil, etc.
If you do find a problem and want some help, give us a shout. Might be best to read-up, start (or join) a thread on this topic in the clock repair section?
All the best!