We have switched from mineral spirits to the tractor supply stuff

I've been down at those tight clearances Brent mentioned - - and I have pulled engines down later and seen some witness marks from the valves and pistons getting together just a "little bit".  Nothing catastrophic - but not very confidence inspiring.  For confidence sake I would add .010" to the low end of those numbers and run it.

Agreed.
You don't need much clearance there.
Them parts aint supposed to move around too far - or you have much bigger problems.

What do the big end and the small end each weigh when on the balancing fixture?

Echoing the crowd.  I use a ton of Powerbond on street applications.  Usually move to the Romac on more serious builds.  ATI for true race cars where fabricated pointers and modified pulleys are pretty much normal anyways.

I posted a comment on the YouTube channel.  He seems like he is kinda lost on that one.
The big rollers are just wrong as a solution to the problem he is having. 

I have built several SOHC engines - not nearly as many as Neighbor or others - and have never had that issue. 
I also suspect that the cams and/or valves are seriously incorrect.

I do think that the elephant's foot adjusters are stupid, and eliminate them by having T&D make rockers with a roller at each end, and adjust the valve lash with selective lash caps. 
Bill C has tested and proven that combination of parts a long time ago.  On an SOHC you can set cold lash on the workbench before assembling the engine.

I degree the cams like two four cylinders - "leading" cam first - second cam after. 
Seems to work for me and addresses any cam to cam chain deflection/stretch - at least cold.
Jay has tested variance in timing as a tuning tool, but I feel you should start at the card specified event numbers and tune from there.

I also toss the front cover snap ring concoction and fabricate a "normal" thrust plate for the stub shaft.
Why on earth would you want a non-confirmable, gasket mounted item pulling on a critical fore and aft alignment assembly. 

Well - that don't look too bad.  Can probably rule out the gasket.
Wish I had better ideas but starting to look like a sh!t happens" sort of deal...