What does this equation mean?

A useful way to see the tradeoff quantitatively is coverage. If a screen library contains g guides and the experimenter wants, on average, c cells carrying each guide to survive Poisson dropout during selection, the number of cells that must be carried through the pipeline scales roughly as $N \gtrsim g \cdot c \cdot e^{\mu}$. where $\mu$ is the expected number of guide integrations per cell under the multiplicity of infection chosen for the transduction (kept low, near 0.3, specifically so that most transduced cells receive at most one guide). Doubling library size to cover more genes, or increasing per-guide coverage to detect weaker effects, both push cell numbers up directly — which is exactly why screen design is a genuine engineering tradeoff between how many genes are tested and how confidently each one is scored, not a free scaling of throughput.