I won't be able to help you, but I join you in your frustration.
While trying to calibrate my 464, I had exactly same feelings and frustration about the instructions (and by just looking to the samples you given, it seems the instructions for the 464 were very similar to yours).
My 464's CRT was clearly worn-out so, at some point I realized I would never be able to get the Fast storage up "to spec" so I eventually gave up. Eventually I adjusted it the best that I could and gave it to a friend of mine.
Like you, I also noticed a gross mismatch between the pre-adjustments that were asked for the Front Mesh and Fast prep, and the final adjustments that can get you to a minimally operating Var Persistence mode.
When you start fiddling with the adjustments... (and only a few of them are relevant to the Var Persistent mode, which limits drastically the number of possible combinations), soon enough you realize that that the only possible settings to those adjustments to get the Var Persistence working at least minimally, completely disagree to the initial settings, so.... couldn't ever understand the rational of it all.
Back then I also asked for advice from anyone in the group who had practical experience on calibrating those beasts, and no heads up...Seem like not much people is familiar with this calibration procedure.
Since my CRT was so worn out, I was never able to tell if all the disagreements I was seeing was because the CRT was way off the specs... or if I just couldn't get the instructions right...
That's not to say your CRT is worn out... I suppose that if yours is so worn-out as mine was, you would know it "on-the-spot". On mine, trace was very dim at 50ns/div, and if I would use at 5ns (x10 engaged), I could only see the trace under complete darkness (and yet, quite dim).
So, if your CRT is not worn out, and you're still seeing the same inconsistencies that I saw during calibration, it probably means that either both our scopes share a similar defect, or that the instructions really don't make sense.
I wish I could help you more (by comparation of results) but I don't have the scope anymore.