There are a LOT of variables that come into play when trying to determine if a server can handle the increase in demand. In this case it seems like your primary concern is the ability of the server to handle the potential increase in mailbox sizes. If you increase the quoto today, in most cases the performance of the server would be the same before and after the increase in quota. Users would actually have to start increasing their mailbox sizes before it would put additional load on server.
Increasing mailbox sizes will increase processing power required when users are accessing their mailboxes, it will increase memory usage, and it will increase disk activity. I would start tracking all of these.
I'd suggest setting up Windows performance monitor to start tracking CPU and Memory. I'd pay special attention to the disks as they are probably going to see the greatest impact and the most likely slow things down . Opinions on what disk queue length can be, vary. I quick google search returned this: "A good rule of thumb is that there should never be more than half the number of spindles in the queue length. If you have a 10-disk RAID volume, the queue length should be less than 5. Such a recommendation calls for capacity headroom to enable peak handlings."
If your disk queues are already near or approaching what you think the limit should be, I would consider upgrading the disks.
If you are storing all data on a single drive, I would consider breaking it up across multiple disks so that mailboxes on are 1 disk or set of disks, the system is on a disk, and potentially even putting the logs on their own disks.
I would avoid RAID 5 arrays. When we see issues with disks not being able to keep up, it typically involves a RAID 5 array.
English
Chinese (Simplified)
French
German
Italian
Japanese
Russian
Spanish
Vietnamese
