At the moment a user simply specifies the overall number of replicas and the default deployment strategy via a statefulset that ensures that pods sequentially scale up and down and as well [I think], the maximum pods down percentage is 25% of the overall total.

So that means that anyone looking at the information provided via the yaml about what the broker requirements are would have the desired pod count, but not be aware that there's an absolute minimum needed for it to function - those could be two different numbers.

There'd be two parts to implementing this; the first is having this specified at all by way of code in the operator as we'd need to check to be sure this API was available [stable in kubernetes 1.21 and I think OCP 4.9 is 1.22 for example], and then optionally expose the configuration through our CRD to the user so that they could control the numbers the operator used when specifying it.