[Guardrails] Scaling down of the cluster shouldn't be allowed post creation of maximum number of buckets allowed

Steps:

1. Install Couchbase server version 7.6.0-1566 on 2 nodes - one node having 8 cores and the other having 4 cores.
2. Initialise a 1 node cluster with a node having 8 cores.
3. According to the bucket guardrail, each bucket would need 0.4 cores, hence maximum buckets we can create are 8 / 0.4 = 20 buckets.
4. Create 20 CouchStore buckets.
5. Rebalance-in the node which has 4 cores. After the completion of the rebalance-in operation, rebalance-out the node which has 8 cores.
6. Now, the cluster has a single node which has 4 cores, so maximum number of buckets per cluster allowed = 10 buckets. But we still have 20 buckets in the cluster.

Hence the issue here is, scaling down of the cluster to nodes having lesser CPU cores shouldn't be allowed.