Great question with three separate answers:
1) Yes, when performing ICGS (unsupervised subtype discovery), ICGS performs gene set enrichment per guide-gene associated cluster using the BioMarker database of assembled marker genes from 300 tissues and cell types. The current database was derived from microarray data. This is a text file that can had additional gene sets added to it by the user (e.g., existing single cell marker genes)
2) LineageProfiler, run either through the command-line or through a workflow will assign tissue or celltype enrichment scores based on k-nearest neighbor classicification. No exact z-score cutoff is recommend (more attention should be paid to the rankings). Again, you can make your own custom database (see our socumentation in our Github wiki).
3) CellHarmony Analysis. This tool is a supervised classification method that automatically assigns aligns cells to existing cell state references from a prior ICGS, Seurat or other matrix of annotated cell state profiles or individual cells from another experiment. See our paper in Developmental Biology from 2017 (Magella et al.) for examples.