This relates to a technology called PERISCOPE (Perturbation effect readout in situ via single cell optical phenotyping) developed by researchers at Broad Institute of MIT and Harvard along with researchers at Calico Life Sciences "that brins the power of microscopy imaging to genome-scale CRISPR screens in a scalable way" (quoting from the article in phys.org) .

They are talking about gene knockouts (possibly related to CRISPR) and then seeing the large scale effects across multiple cells. Unless, there is some new method / algorithm developed to handle the combinatorial explosion , i do not see how the above could be harnessed practically.

Ok, possibly they could possibly have an idea of the heiracrchical network of transcriptional factors, possibly obtained during embryogenesis / morphogenesis and some idea of the coregulated sets of genes in various tissues.

Possibly, a better method would be to silence the transcription of all the genes (not knockout which wud be permanant) and switch them one by one starting with the top of the heirarchy.

Of course, everything would depend on the methods available for analysis