When I look at the first table I see a number of 220k non Covid death overall compared to 60k after the 2nd dose. Assuming those non Covid cases are constant over time I would assume the sample size is still bigger than 1/4. while not perfect that doesn’t look like „not enough samples“ to me.