I'm not convinced you are, at all (other than perhaps, very slightly and very pedantically, in the sense that some purely-price-action parameters can, in fact, be fully automated, and in that sense therefore be said to involve no "discretion" at all ... but I doubt very much whether a retail trader could realisitcally build a viable, steadily profitable system just around those, so it's kind of "academic", really.)
Other than that minor, parenthetical, academic point, I agree completely with the broad thrust what you've said, anyway.
For what it's worth, if anything, I very much share your perception that given your background, circumstances and objectives, price action (i.e. rather than indicators) is very likely to be "the right way to go". I make this comment perhaps with some bias, since it's how I trade, myself, and I offer it in the knowledge that not everyone here will necessarily agree with it.
I also strongly agree with your sentiment that one must be consistent if wanting not to screw up one's odds in the long run: long-term, successful trading isn't necessarily about doing anything difficult - it can be about doing perfectly simple things, but very reliably and very consistently. But the process of learning how to identify and execute them is still in itself necessarily difficult and time-consuming (in my opinion, natch.) To put it bluntly, price action is going to be needed to "get you there" (with or without indicators, actually), but admittedly it's harder and slower to learn than indicators.
I suppose, very pedantically, one could try to "debate" that and produce some kind of rarified counterexample, but for all practical, reasonable, relevant purposes I agree with you, anyway.
It would be really nitpicky (even by my standards!) to dispute what you've said above. My perspective is the same as John's (just above): price action is (realistically) "discretionary" ... and indicators can/should be, too, really.