I don't care much for the search functions in slope/w, but they are better than nothing.

any of them?  or just the auto-locate?  Is there any particular reasoning?

I like the auto-locate feature.  I have found it finds the critical surface in most typical analyses I have performed, and without the extra work of specifying limits or other parameters.  However, if you are looking for a specific surface that is unusual, some of the other search options allow you to better define the search.

Just don't trust the search too much.  Just Saturday afternoon I was reviewing a report containing some Slope/w results where the program had searched out nonsensical failure surfaces that gave lower factors of safety.  This was due, I think, to some negative stresses created by the goofy geometry.  (I've seen a similar thing before, but this time I only had the output sections, not the file, so couldn't verify.)  I don't think I saw those so much from UTEXAS2 back in the olden days.

The search routines in SLOPE/W are too dependent on the inputs for the parameters of the search.  So if you are not careful when you set up the search or run additional searches with other inputs, you often miss the critical failure surface.

The UTEXAS routines were much more robust and would continue to search to the local minimum.  Maybe it's because UTEXAS is what I really got started on, but I just like it better.

No search is perfect and you have to LOOK at the results to know what you got.