Fix mistake on ResourceModel page

[RainbowMango]

What would you like to be added:
Fix the mistake on ResourceModel page as discussed at #713 (comment)

• en doc: docs/userguide/scheduling/cluster-resources.md
• cn doc: i18n/zh/docusaurus-plugin-content-docs/current/userguide/scheduling/cluster-resources.md
• code: (@RainbowMango, Fixes incorrect resource model selection karmada#5706)

Why is this needed:

[LavredisG]

Opened this: #736, check it out for any suggestions/corrections etc.

Also, I think in this line, this part is not correct "and the nodes of model2 and later as meeting the requirements.", according to our discussion, since memory request is not met by grade 2 nodes.

[LavredisG]

I would also like to ask another question on the topic. Suppose that we want to schedule for example 3 Pod replicas. We get that cluster1 can fit 6 and that cluster2 can fit 5. In this case, cluster1 is chosen based on cluster resource modeling, but will it schedule all 3 of them at the same time? Or will it pick them serially, because in the first case we will get {cluster1: 3, cluster2: 0}, while in the second we will get {cluster1: 2, cluster2: 0}. Examples were based on 1-pod scheduling, so I couldn't figure what happens in that case.

[RainbowMango]

Suppose that we want to schedule for example 3 Pod replicas. We get that cluster1 can fit 6 and that cluster2 can fit 5. In this case, cluster1 is chosen based on cluster resource modeling, but will it schedule all 3 of them at the same time?

Actually It depends on the scheduling policy. For example when dividing replicas as per StaticWeight, like 2:1, then cluster1 will have 2, and cluster2 will have 1.

I think the resource model is only used to figure out available replicas of each cluster, it doesn't control how to divide them. I also mentioned on #736 (comment).