Thanks for your questions; they're insightful. I have a few concerns about releasing the limit setting for an application:

1. It might consume more resources than expected, especially with a large team that may struggle to understand and collaborate closely with app development.

2. Resource contention could lead to excessive resource usage, leaving other critical applications underutilized, which is not acceptable.

3. Quality of Service (QoS) will be burstable, not guaranteed, and may be preempted before guaranteed pods in case of node resource pressure.

4. Due to the QoS policy, the application can't benefit from static CPU pinning.

5. Ensuring two replicas don't run on the same node is crucial to prevent resource competition and unexpected outcomes.

Consideration should be given to a dedicated node style, but it's essential to evaluate if the benefits outweigh the associated costs.