Do you actually NEED to calculate multiplicity of infection (MOI) when establishing a stable cell line using lentiviral transduction?

I will offer 1,000 RSC to the best answer to this question, and an additional 250 RSC for up to 2 other answers which are useful for informing my experimental design. 

I'm going to be using 3rd generation lentiviral transduction to take a cell line which has a knockout for a protein and create clones which express wild-type and various mutants of this particular protein. The vector I've cloned my gene of interest into has GFP, so my plan is to transfect HEK293T cells, harvest viral particles, transduce my target cells, use FACS to sort GFP-positive cells, do a limiting dilution, and then grow up clones. I'm thinking of using this protocol (https://rnai.genmed.sinica.edu.tw/file/protocol/4_3_EstimationLentivirusTiterGFPV1.pdf)

Is it okay if I never measure the actual number of viral particles? I just want to transduce in such a way that each cell gets infected once, so I'm thinking of transducing cells with a variety of dilutions of my viral particles, and go with whatever dilution of the particles gives me less than ~20% of cells fluorescing, which should be each of those cells infected roughly once. 

Am I missing something here? Why would you ever need to measure MOI? To me it seems superfluous to measure the number of actual viral particles when the number you need is also variable depending on cell type. Never done this before so I'd appreciate any insight!