If I understand the artemis thing correctly, the heat shield still performed more than well enough and none of what they’ve seen presents any danger itself- basically, charred bits that were supposed to stay on, fell off, but they’d already done their job and didn’t hurt the performance at all or present a risk.
But, just the fact that it didn’t perform as modelled is a worry for them because it undermines the model as a whole, so even a no-risk deviation raises some concerns about the possibility of other deviations. So now they’ve got to figure out why it deviated and basically update the model so that it exactly fits the reality, and then see if there’s any other concerns. Which basically would leave the question, is there other stuff like this that we just haven’t see yet.
It seems a bit like the Starship rocket test- the number of failed rockets was within the safety margin, but at the same time any sort of failure is a concern because the same thing could happen more, and equally any failure kind of goes to the overall trustworthiness of the vehicle, because while it’s OK that rockets failed, they shouldn’t have.
That could be oversimple though.