I'll leave the discussion of the machine to Deejay, who is our reel-mower guru, and just comment on the welding.
As you will find if you read the thread I nominated in my previous post, there are two main issues in the welding project. The first is, ending up with the mower frame correctly aligned rather than corkscrew-shaped. This is by far the more difficult part. The second issue is making the machine strong enough to prevent another failure.
The alignment issue was handled in some detail in the other thread. Essentially, you need to clamp the frame into correct alignment, tack it, unclamp it, realign it by twisting, re-clamp it, weld it a bit more using short runs which compensate for each other in the distortion they cause, unclamp and align it again, and so forth. The core technology here is to carefully put each small additional weld in a position where the shrinkage distortion will twist it in the opposite direction from the previous small weld. Of course it becomes more and more difficult to correct the twist by physical force as you put more welds into place, and you have to do it by putting the welds closer together on one side than the other, so there is more total shrinkage force in the required direction. While this kind of thing is fun, it is very frustrating to have to keep grinding off welds after you make mistakes that cause distortion. A genuine tradesman welder (I do not mean an experienced welder, I mean someone who completed the welding apprenticeship then practised the art for a decade or so) can do this kind of thing routinely without breaking into curses and starting again repeatedly.
The other issue is to end up with the machine strong enough to not break again. Presumably you will be using a grub-screw type engine-side clutch-half next time around, so the strength of the original SB design might actually be sufficient, but if you just weld a reinforcement underneath the original rail you probably won't even achieve that. The reason the original rails broke is that the vertical flanges on the rails were neither deep enough nor thick enough to adequately control the vibration-induced deflections involved. It is difficult to change this situation without making new rails, of deeper section and thicker steel. If you did this by laying the new rails inside the old ones it would not only look pretty awful, it would only be possible to weld the new rails to the side plates on one side. A series of spaced stitch-welds on both sides of the rails would give you a stronger result together with a better opportunity to control the shrinkage distortion.
