KEGG passway analysis after Seurat analysis

Hello, everyone. I am so sorry for this amateur question.
I have a scRNA-seq data and want to compare and visualize the expression levels across genes on a cluster by using clusterprofiler package (passway analysis), but I think the average expression levels of each gene can not be compared because these are not the same scale.
So I tried 3 steps as follows.
First, I selected a passway that I wanted to use.
Second, I calculated average expression of each cluster for each gene.
Third, I scaled them for each gene.
I think this is as same as Seurat DotPlot function, however, my method does not reflect the percentage of cells expressing the genes.

My question is

1. Does this method valid? or has risks?

2. Is logFC better than other transformations to try passway analysis when we use Seurat results?

Thank you.