One way to get more results for a surname project is to extract them from the SMGF database. SMGF has stated they they do not mind if people use the results for their surname projects.
Lack of the donor's identity is not too important, since you are really interested in the ancestors' haplotypes. They can be identified from the pedigrees. Identifying the ancestral, or patriarchal, hyplotype is usually the primary goal for a family genealogy DNA study. It may turn out that one of the persons needed is already in the database.
Simply do a surname search. Any haplotype will bring up those with the surname in the database. However, it is best to try it with modal haplotypes of the various haplogroups, since they will give closer matches.
If you don't already have a predicted or tested haplogroup, try the R1b modal first. Most people in the database are in that group and their haplotypes will be closest to it.
Also try R1a and the I modals until you get the closest match. I haven't had to use others. However, these may be close enough to others for this purpose.
You should retain the modals for future searches by bookmarking the results page. You will have to change the name of the page with your bookmark editor. Otherwise, they will all end up with the same name.
After you have found the the closest matches for each entry, increment the appropriate modal values of the non-matching results by 1. Do this for all non-matching results at once. Otherwise, you will exceed your daily search limit.
Next, decrement the remaining non-matching results by 2 (-1 from the modal) and search again. Then increase the non-matching results by 3 (+2 from the modal)and so forth until all the results match.
Mismatches on duplicates are much harder to extract, since you have to identify both values at the same time. It helps to use the frequency tables at: