I'll recommend Imago training [1] to any and all couples. It's reflective listening on steroids and was transformational for us.
The basic process starts w/reflective listening, then goes deeper to get at the underlying wound ("when you do that, it brings up all the times my parents..."), then to what might happen in a perfect world ("the toothpaste cap would magically fly back on the tube after 30 seconds of inactivity"), then to some concrete make-ups.
The other things we do that helps is to stick with the current argument (which I find difficult, sometimes) and to not go "meta" ("see, honey, there's a pattern here where you ...").
It's often worse than "meta" though, where a normally reasonable person will intentionally (though probably subconsciously) confuse and mislead when in the heat of an adrenalized fight or flight situation.
I'm not sure why you're being downvoted - that reads as a pretty fair question to me. I can't tell what makes the meta argument meta. But I can tell why one is going to lead to a much worse argument than the other:
Saying "when you do that, it brings up all the times my parents..." suggests that the problem lies in what the speaker is thinking and the speaker clearly knows it.
Compared to "see, honey, there's a pattern here where you ..." - the problem is still in what the speaker is thinking, but that sentence suggests the speaker doesn't understand that and is about to present their thinking as historical fact. If they phrased it "honey, what you just did made me think there's a pattern here where you..." would probably still be a bruising conversation but it will probably end in a much better place. The speaker has, based on a lot of arguments I've seen, almost certainly misunderstood the pattern they think they see.
The basic process starts w/reflective listening, then goes deeper to get at the underlying wound ("when you do that, it brings up all the times my parents..."), then to what might happen in a perfect world ("the toothpaste cap would magically fly back on the tube after 30 seconds of inactivity"), then to some concrete make-ups.
The other things we do that helps is to stick with the current argument (which I find difficult, sometimes) and to not go "meta" ("see, honey, there's a pattern here where you ...").
[1] https://en.wikipedia.org/wiki/Imago_therapy