By surface features I mean mainly dark spots analogous to sunspots, or starpatches showing unusual velocity fields, or chemical areas. Small scale features like granulation are treated separately. Although we see the door opening slightly to direct resolution of stellar surfaces with large-aperture telescopes and particularly interferometers, most surface features are detected and measured using spectroscopic analysis. The distribution of Doppler shifts arising from the rotation of the star is a one dimensional map of the star's surface. Therefore, whenever the broadening of the spectral lines is dominated by rotation, we have the ability to resolve the surface, and to see features if they are there. Sufficiently high rotation for cool stars occurs only in tidally locked binaries. When the rotation is small, i.e., normal for cool single stars, the surface cannot be resolved, but 'spot' information can still be deduced from the variations of line bisectors and from photometric variations.
The Vienna meeting (held 9-13 October 1995), International Astronomical Union Symposium 176, Stellar Surface Structure, dealt with this topic (1996, Kluwer: Dordrecht, K.G. Strassmeier & J.L. Linsky, eds.).
Some work along these lines can be seen in these references: