Better ACF pValues and use pVal not 1-pVal
Add the new spfun/studentT.ccPv
for exact pValues(autocorrs) rather than approximations that sometimes do not even converge asymptotically. Since this can get slow for many thousands of points, keep old formula as a default for now.