A Head of Product can't review every accessibility decision and I don't think they should even try.
I bet this makes you feel like you're off the hook. It may even feel reassuring at first. But mind you, all important decisions still get your attention if you want to catch mistakes before they create rework.
You just don't want to become the bottleneck. If your people start waiting for approval instead of using their own judgement, accessibility will either stall or happen without enough confidence.
I think a much more appropriate goal is to help teams make sound decisions when you're not in the room, as if you were.
And that only works when people know what good looks like. They have clear expectations for each feature and someone to ask when the answer isn't obvious. You can focus on repeated patterns and risks and less on the day to day.
You can take a step back. Just be careful because stepping back can also fail.
If expectations start to become too vague, people will fill the gaps differently. If raising an accessibility concern always puts a deadline at risk, people may stay quiet. If nobody has time to help, "the team owns accessibility" becomes another way of leaving them alone with the problem.
Autonomy is important and autonomy needs support more than anything.
Give people enough clarity to decide, enough safety to ask and enough time to act.