You ask "What's going on with Monterey? "
What were the average speeds achieved with your previous macOS, without that as a base-line we have no point of reference.
You also state "Swapping cables does not change the outcome."
What standard and specification patch cables are you using, what and where are you swapping cables?
As a general rule, the maximum network speed achieved is dictated by the number bad packets detected by your router, switch and by the computers NIC.
The network speed is auto-adjusted by all the devices in the chain.
The router will auto-adjust its upload/download speed to minimise bad packets and packet collisions between the switch and the router while at the same time the switch will auto-adjust its upload/download speed to the computers NIC to do the same.
This is a dynamic process and is not fixed.
The maximum speeds that are quoted for any Router, Switch or NIC are never obtained in practice, these are the theoretical maximum speeds obtained when the patch cables are short and run in straight lines, are impedance matched and individually tuned by trimming the length to minimise SWR, are double screened and solid cored, and finally all the connected equipment must be powered from the same single-point mains outlet or all on ground isolated power supplies to eliminate ground-loop noise on the LAN cables.
The only certified way to test a network for speed problems is with a tool designed for the task, a Network Copper & Fibre Transmission Bandwidth Tester, which is used in various locations throughout the network to pinpoint where problems are occurring.
Without a bandwidth tester to perform quantitive measurements everything else is just guess work.
Since network Bandwidth Testers cost around $1,500 I'm guessing you don't have one, or have access to one?
I would suggest that the only simple test you could make is to connect the iMac Pro direct to the fibre router using a short double-screened CAT6 solid-cored patch cable and without any bends in the run.
Each bend in the patch cable reduces transmission speed and if the bend radius is tight enough then speed may be reduced by more than 50%. The cable manufacturers specification sheet will state what the minimum bend radius should be and what reduction in transmission is expected when that bend radius is used in a typical LAN installation.
Measure the speed by aggregating the results from several different on-line speed test sources and test when network traffic is low, late-night or early-morning is best when businesses are closed and people are not consuming bandwidth by watching streaming services etc.
When speed-testing using on-line sources be aware that the speed quoted is between you and wherever that on-line source server is located, use a server look-up service to find a source close to you, testing speeds using a speed-test server on another continent will yield meaningless results.
If you find that the direct connection between the iMac Pro and fibre router using a short - straight - solid-cored and double - screened CAT6 patch cable gives vastly better speeds then re-insert the switch and use a similar high-quality patch cable between the switch and the iMac Pro and measure again.
By a process of elimination you may be able to find the pinch-point on your network that is throttling the transmission speed.
If you are using cheap, unscreened, CAT5e or CAT6 flexible-cored patch cables between your fibre router and switch and between switch and end devices then this will most likely be where the problem originates and nothing at all to do with "Monterey" as in the title of your question.
If you are unable to resolve the problem of slow speeds to your Apple devices using the above suggestions then arrange for a network specialist to carry out a Bandwidth Test on your in-house LAN, they will be able to give you certified report of the performance of your entire network and identify where any pinch-points exist.
If relying solely on on-line speed-tests to determine the network performance of your in-house network and devices remember that professional network engineers do not use on-line speed testing services for anything other than a qualitative or comparison quick-test. On-line speed test services are not suitable for making absolute measurements since you have no way of determining the performance of the remote server that is providing the service.
HTH
Will.
