Why are they lumped together? Who has it wrong? It's not the fault of FamilySearch. It is the fault of Adelbert Yoe, the enumerator.
Yoe apparently got off on his numbers. The entry above linked me to an image of the 1920 census which I had actually already located (1920 U. S. Census, Hancock County, Illinois, St. Albans Township, ED 25, sheet 7A, Charlie Neill, household, lines 42-47). That image of the actual census is shown below. To be honest, I had never noticed the previous household headed by Elbert Young had the same dwelling and family numbers as the household headed by Charles T. Neill.
The lumping together appears to have been a simple error on the part of the census taker. If the households had been actually living together in the same dwelling, the dwelling number would have been the same and the family numbers would have been different--and it's pretty clear that these are two separate families.
My citation for this entry includes the line numbers because the line numbers are unique to this household--the other numbers are not. My analysis of this entry should include a brief discussion of how the dwelling and household numbers are repeated for two families in a row and why I think this is an error.
I'm a strong believer in citations and in my work (and in Casefile Clues) I cite material in the spirit of Evidence Explained. Here on the Rootdig blog, I have a different philosophy. Posts made here have enough information that the reader could locate where the material was obtained.