In phospho/pan analysis in Western blots, what is best way to normalise to an internal loading control?

I am analysing the expression of a protein kinase X that is a phosphoprotein through Western blots. I have labelled the membrane for both the phosphorylated form of the protein and also for the total target protein. I am comparing the expression of my phosphorylated protein between different experimental conditions.

For each sample, I have calculated the proportion of protein that is phosphorylated by dividing the phosphorylated protein signal by the total target protein signal. I have also done this to ensure that the differences in phosphorylated protein between samples is not due to variations in total protein levels under different experimental conditions.

However, I want to normalise this ratio to an internal loading control, since I think that it would be incorrect to assume that the expression of the total target protein is uniform between the different experimental conditions. I have read when comparing the expression of a protein amongst different experimental conditions, you need to use an internal loading control.

I have used a total protein stain (Ponceau S) instead of a housekeeping protein as my internal loading control, since studies have shown that expression of housekeeping proteins can vary between different experimental conditions and that the level of total protein between different experimental conditions is more consistent.

I am currently debating between the following two methods of normalisation:

(1) (phosphorylated protein X/Ponceau)/(total protein X/Ponceau)

(2) ((phosphorylated protein X)/(total protein X))/Ponceau

I think that method (1) is the more correct way to normalise the phospho/total protein X ratio, since you would want to obtain the relative abundance (to Ponceau) of both phosphorylated protein X and total protein X before calculating the phospho/total protein X ratio.

However, I would was wondering if anyone knows whether this reasoning is correct? Any insights are appreciated.