I should add that I had the 46-marker test, as did the 8 near matches which I mentioned.
I have a similar problem. My nearest 250 odd y-dna matches consist of about 200 different surnames. Only 1 has any similarity to my own (Milloy versus Melloy), and that has a MRCA of 15 generations. I've read all I can about non-paternal events, convergence, links going back before the use of surnames etc but there appears to be no rhyme or reason at all to the list of names. Even the 8 nearest matches (with an MRCA of 6 generations) consist of 5 quite different surnames. What on earth is going on?
Most of the Molloys, Malloys etc in the database have round about 20 to 30 generations to MRCA. Should I ignore the close "matches" and work with the Molloys, or should I ignore the Molloys and assume non-paternal events have affected my nearest matches????