I have contacted two businesses both do headstones and table tops. But they don't have information on how true the tops actually are.
One is local and I'll visit it if I get from my work any time when they are open.
I have tried two method home:
1) Support good straight edge on top of the 0,05 mm feeler blade from both ends. Straight sample should allow a same gauge feeler gauge to slide under the straight edge. I found that same gauge feeler blade is just a tad tight....as it should be. I have 0,03 0,04 0,05 0,06 and 0,07 mm feeler blades. This shows pretty easy if there are any big problems.
On real surface plate 0,04 mm feler blade slid sweet, I could feel a little constant resistance. Same thickness feeler gauge was tight and tend to move the straight edge...pretty sure the DIN 874/2 straight edege has some 20 um variation....
On granite kitchen top piece 0,03 was tight on some spots and couple spots could place even 0,06 mm feeler gauge under the straight edge that was supported 0,05 mm feeler gauge at both ends, but it was nearly as bad as I tought.
Hair staright edge did emphasis high spots and walleys pretty clearly. Looks like colorful table top granite is not really a best choise. It might have harder and softer parts in it and it seems to show on surface quality.
2) I place straight edge on sample plate and imobilize it with a long parallels on both sides, Then I slide a magnetic TDI stand w/ clock on the side and indicate from top of the straight edge.
This was interesting...Real surface plate showed practically no clear movement on the dial...which was good, because it showed that the measurement can be used to indicate straight line and instrument was good for it.
On the othe had the checkered granite surface plate showed regular waviness and raising/dropping error trough the whole travel. Indicator base is relatively short, it would pick up easily surface undulation. Another thing is that DTI was at offset from the base amplifying the error greatly. I did not do math, but on real surface plate the indicator hardly moved at all and my sample of granite table top indicated worse than 0,05 reading, growing to 0,1 mm at the ends and also showed definite dip on the middle of the plate.
Now I have a way of field checking the embryo surface plate in a somewhat reliable way.
VT's comment on using a building aluminium straight edge is haunting me. I know that there are some very good aluminium straight edges, but I can't get my brain around the oxide layer/truing abrasion when used against feelergauge atc problems. But it could be cheap, rigid and lightweight. Can't really protect the soft edge with a steel strip from feeler gauge roll.
Took two solid hours to get repeatable results, but I think I have a method for testing now.
Pekka