Grouping everyone with an unknown father under one "unknown" person
I'm wondering if this is considered to be an acceptable way of grouping people (it seems strange to me): someone researching Borzecin (Poland) is gathering up everyone with the same surname (for example Pudelko) whose father is unknown, and putting them under one "unknown Pudelko" man as his children. This one (GKPG-163) has 51 so-called children, with birth years spanning about 100 years. Wouldn't this mess up how the FamilySearch algorithm works in giving hints? I'm considering unlinking these 51 people, but I don't want to interfere with a system that might be the newest thing. What do you think?