If you have ever spent time with test boring crews, you will understand how crude the test is. In the old days with rope and cat-head, all sorts of problems can occur. With mechanical hammers, lubrication is an issue. How about inertia in a string of rods? Then, some day take a look at the split spoons. Buggered up ends are possible. All this will tell you that precision just is not there. If you want to work to a few decimal points with N values, go ahead. Not me.
I agree with you that SPT is a very crude test. However, SPT correction has been used everywhere else to correlate to density, strength parameters, etc. Why not Site Class?
seed and idriss correlations for liquifaction are based on corrected N-values. I am not familar with the cited method and rarely do such work. Just saying, the source of n-value correlations uses corrections.
Agree with moe...N60. If you are thinking there should be a bump in site class, move toward shear wave velocity assessment. The SPT approach was built in conservative for many of the reasons mentioned here. If N60ave is better than about 30bpf, there is a good chance the shear wave velocity will get you to a C but depending on geology and what you're beating the pipe (ie. Spoon) through along with who is doing the spoon driving and type of hammer.
I think that the use of SPT N-values (corrected or not) for site classification purposes, will limit the site class to "D". For example, if you want to classify the site as "C" you will need an average SPT N-value of more than 50 for the top 100 feet (30 meters) of soils, which is rarely encountered. If there are cost savings when classifying the site better than "D" and knowing that your geology may allow it, perhaps a shear wave velocity method to classify the site is more appropriate.
its not a good idea using SPT on soft rocks . These materials have their own behavior which deviate from the well known soil behavior or the rock mechanics , be very careful about what you are doing