Thomas, that is what you have to use.
But it is not automatic.
When the saddle is fitted to the lathe ways, the surfaces of the saddle that match the lathe ways should be parallel to the (under) surface to which the apron mounts (so everything that connects to the leadscrew stays parallel with it) and while there is no particular cutting constraint, the scraping the saddle ways should not cause the apron to tip out or in too much (maintain apron more or less square vertically). IOW, the when complete the saddle should lay on the ways such that the apron mounts back up aligned "pretty well" vertically and horizontally with the leadscrew; and does not hang in or out too much. Scraping "straight down" will tend to keep you on track well enough, but check every now and then.
The cross slide ways ultimately need to be slightly out of square such that the cross slide will face slightly hollow, going from front to center or from center out to front (the "normal" direction). Meaning that if a tool were placed to feed from the back of the lathe to center, or center to back, the face would be conic. This can be adjusted when the cross slide ways on top of the saddle are scraped for the cross slide. But worth being aware which direction they are trending, when scraping the underside to the lathe ways so long as the more critical apron geometry is maintained.
The top of the cross slide should be parallel with the lathe ways, so that the compound does not point up or down with different swivel locations. But this is also not terribly sensitive, just something to be aware of and check. Of course it could be adjusted after fitting the cross slide, too.
The headstock does not have a lot of surface under it, and can diverge quickly, so be careful. The headstock should point straight down the ways, not to either side, certainly not down, and only enough up to be certain it is not pointing down.
The TS is the same idea. It must not point to either side as it is extended, and should point up maybe a couple thousandths in the full travel, to give some durability before wear causes it to droop. (this is off the top of my head to describe the general geometry. If you can check Connelly, Schlesinger standards, or SB's own specs for TS pointing up, it would be best)
http://totallyscrewedmachineshop.com/documents/Testing Machine Tools (Dr.Schlesinger).pdf
If you don't build a new base for the TS, it will probably need to be shimmed to get the center back up to headstock center height. Or the headstock could be scraped to lower it, but I would not do that. It is too easy to go wrong for minimal benefit. The headstock does not get any wear under it and should not need much scraping other than to get the angles to fit the new ways. I have mentioned in the past that my approach is scrape one surface only, complete, before moving to another, under the headstock, when fitting it to new ways. That way if it is parallel, it will stay parallel. If the headstock does happen to point up or down, then this rule will have to be modified to convince it to tilt the other direction; but make a sketch and think it through.
smt