If we wanted to study the structure of a plasma membrane, why are red blood cells a more attractive cell type to work with than other cell types such as liver cells or kidney cells?
Human RBCs are relatively simple in structure compared to the other cells in question; they contain no cell organelles (so far I knew; please correct me if there is any new theory) and therefore contain only one membrane.
RBCs do not have significant amount of extracellular matrix making them very easy to work only with the cell membrane. RBCs float in a fluidy medium (blood-plasma), so easy to collect, distribute in containers, keep into various solution. no need of maceration. Human RBCs are very uniform in shape and size, and do not divide. That makes it useful for various experiments like demonstrating plasmolysis and deplasmolysis as well as quantitative experiments on membrane biomolecules.
One classic example is Edwin Gorter and F. Grendel that first showed that cell membrane is bilayer; that is upon observation on RBC. They took known number of RBC, denatured the membrane, extracted the lipids and spread that as a lipid monolayer on water-air interface. They found that the monolayer is twice bigger than total surface area of taken RBC !
*The Cell/ Cooper/ 4th Edition/ASM-Press and Sinauer publication.
However this is not a reference-based answer, but these are the features more than enough to make me so lazy that I would choose only RBC and flyaway from any-other option.