I was not paying attention to the fact you were looking thru all 3 stages.
The stability problem, if there is no fdbk from any stage back to another, is within the stage.
Unless of course there is some parasitic fdbk of significance you are unaware of..
So stability here, in general for cascaded stages, with no interstage coupling, overall phase is not of concern
as long as each stage has adequate phase margin. But phase, if you are seeking a specific phase response, like
linear phase, is impacted by each stages response (pole locations, gain). Whole different problem.