Some thought about a proxy server problem from textbook

- Description:

Assume the rate of the institutional network is Rl and that of the bottleneck link is Rb. Suppose there are N clients requesting a file of size L with HTTP at the same time. For what values of Rl would the file transfer takes less time when a proxy is installed at the institutional network? --(Assume the RTT between a client and any other host in the institutional network is negligible.)

---

I understand that the goal of setting up a proxy server is to satisfy client request without involving origin server, and it is a way to reduce response time for client request and reduce traffic on institution's access link.

But I am wondering why it asks about institution bandwidth? Is there any limit to that? (isn't it the higher the better). Doesn't "cache hit rate (requests satisfied at cache/ all requests) > 0" already guarantee less response time?

thanks for any idea!