Sometimes it may be better if the code does take up more than one screen!
This is the most densely packed code I've seen.
And what is the purpose of feedback nodes on error clusters that aren't even part of a loop? It makes it hard to tell which is the first function that executes. Is that so you can hold on to errors from the last time you ran the program? "I had an error and stopped the program, but everytime I restart it won't work and I get the same error! I can't figure out why!!!"
Most likely the author of this code wire things randomly together, accidentally creating loops and these feedback nodes were inserted automatically (the "auto inser feedback nodes in cycles" option is enabled by default). For example you can delete the rightmost feedback node and simply wire the remaining ends together and the feedback node will magically reappear.
I wonder why all calculations are done in DBL, but displayed as SGL. 😮
You're right that is scary. And worse. I especially like the amount of overlapping and parallel wires in the right hand side of the window of the top picture.
I was impressed with how in my example, the programmer worked hard to really compact everything, and by doing so was able to squeeze it into much less than one screen, so it was like they were trying to follow one style rule!