Calcium isn't nearly as mobile as nutrients like potassium so this helps with surface applied lime to keep it closer to the top.  I only have experience with 6" samples and 4" samples.  6" is too deep for no-till.  2" is too shallow for P and K levels.  I am familiar with the acid zone and UK did some work on this some years ago.  They found the 4" level to be the best.  There is a zone that becomes more acidic, but surface applied lime helps with this and will stratify over time.  I suppose that if the root zone were only 2" deep it would be more critical, and there may be reasons that some soil types and some regions of the country might be different and there would be merit to pH testing more regularly the top 2".

I might do some research on the "acid roof" to see if there is new data worth considering in managing no-till.  One potential problem with applying 1 ton per acre is the cost per ton goes up because applicators still have to drive over the same numbers of acres while putting on less lime.  I do like to grid spread lime if fields are variable.  Once they get adjusted, I don't see them vary with crop removal as much as P and K based on soil tests at this point.

As for grid sampling, I have moved toward fertilizer applications based on crop removal once I have test levels at a satisfactory level.  I started with 2.5 acres grids by soil type, but I am changing to grids based on soil type and yield management zones.  This will mean larger grids that match the yield monitor and soil type more than looking like a checker board of 2-2.5 acres.

I may still use soil samples to in some areas to try to determine problems or fix problems, but I am becoming a stronger believer in the zone management theory over the grid management theory.