Anyone else finding "AI Overview" targeting completely different from classic featured snippet optim

Been running SEO for a B2B manufacturing client (niche industrial equipment, low search volume per keyword but high commercial intent) and noticed Google's AI Overviews behave pretty differently from how featured snippets used to work.

A few observations from testing:

• Straightforward "what is X" definitional content still gets pulled the way you'd expect, similar to old snippet logic.
• But for comparison-style queries (type A vs type B), AI Overview seems to synthesize across multiple sources rather than pulling from a single "winning" page, which makes it harder to reverse-engineer what specifically triggered inclusion.
• Pages with clear FAQ schema still seem to have an edge, but it's not as deterministic as it used to be for snippets.
• Content freshness seems to matter more than I expected — noticed inclusion shift after updating publish dates and refreshing stats, even without major content changes.

Curious if others working in niche/low-volume B2B spaces are seeing similar patterns, or if this is mostly a high-volume consumer search phenomenon. Also wondering if anyone's found a reliable way to track AI Overview appearances at scale beyond manually checking SERPs — the tools I've tried so far (Semrush, Ahrefs) seem to lag on this specific feature.