lewm
"Just one teeny, weeny thing: Back up the thread, Elliott recommended to set AS to equal VTF or some high percentage of VTF."
lewm,
either someone else, or maybe I worded something wrong,
but I use the Blank LP method, don’t even think about VTF, just start AS at zero, spin platter manually (easy with direct drive) watch the arm pull in (if it doesn’t there is something wrong), add anti-skate a bit at a time, checking in the two approximate null point locations, and set it with no outward force, consistently floating, or if needed to avoid ’out’ set it for a speck of inward pull, rather than a spec of out.
I don't look at VTA before or after, just the actual results.
I re-check perhaps monthly, to confirm both VTF, AS, and if removable headshell: view Azimuth from front with mirror method.