The combination is bad.

I'm not really sure what there is to do about that, then. My own project is already is about to hit 3 years on something that was intended to be <1 year total, due to constant scope creep. Nothing bad seems to ever come out of the delays though, so I tend to ignore most of the complaints.

If you see it as an argument

I don't really see it as that. "Discussion" is more what I try to do. But you are correct that I don't think I can argue on their terms.

are you sure you understand what they value and prioritize

Probably not exactly, but my point is that the priorities technical leadership says we value (quality, scalability, fast iterations), run counter to what we actually prioritize. I often ask why we prioritize Project X over Project Y and the answer is almost always a variation of:

• "We can't let IT be the reason the Project X is late."
• "The business thinks we've been working on Project X a long time (often not true) so we need to show progress."
• "Project X was promised for Release Z so it needs to get done over anything else."

Which is why I said our priorities are more about appearing busy and important than anything else. (My own project isn't even wanted by most business users. It was spearheaded by the VP of IT as a huge technical modernization effort despite doing almost nothing to improve or get away from the legacy system it is "replacing".) So I think the reason I have such trouble getting buy-in is that better testing runs counter to IT's true priorities, even if it provides business value.

[Trust] might be eroded down due to the consistent failure to meet estimates.

Perhaps. But trust is already pretty darn low for that very reason.