In a hive of 2 deeps, or mediums, there is sufficient storage space to contain enough honey and pollen to see a hive from late September through early March, especially if they've been allowed to build burr comb after honey harvest. Feeding is simply an insurance policy to avoid deadouts as a result of rapid brood development if an extended period of adverse weather should crop up, which it often does.
The queen will cluster in the brood chamber. She is very unlikely to cluster on storages combs that have never had eggs laid in it. The worker bees will work those areas but the queen never goes there unless it is to escape a danger and she will only turn true storage combs into brood combs if she is somehow stuck there. What that boils down to is that a queen won't move the cluster into a super of nothing but stores, the cluster will remain in some area of the brood chamber and the worker bees will fetch and ferry the needed stores from the storage combs to combs within the cluster area.
I've observed the following behavior of bees in cluster:
1. The cluster, if initiated in the lower brood box, will move to the top of the brood chamber (top of upper brood chamber) consuming the stores as it goes. In this case the stores in the outside storage combs might not be touched (especially in the lower box).
2. If the cluster is initiated in the upper portion of the brood chamber (upper box) then the workers break cluster, when temps allow, and fetch and ferry stores which are deposited within those cells within the cluster. In this case the cluster remains stationary the entire winter.
3. If a box of stores, never having any brood reared thereon, is placed above the brood chamber, the cluster will stop their upward movement through the stores and resort to fetch and ferry.
If a queen excluder is placed between the brood chamber and a super of stores then the action of the bees is as in case #3. However, the queen excluder can create a situation of stores not being accessable to the bees due to the fact that the metal used in the excluder will draw the cold creating an additional barrier besides the grate of the excluder. In an apiary when ecluders are used extensively the use of the excluder might not be a problem, but bees are creatures of habit, but in an apiary where excluders are seldom, if ever, used, using them in such a way is most likely a death sentence.