We can and we do already have multiple THORs running on the same physical h/w - they are independent / unaware of each other. They run jobs as they receive them, regardless of what the other is doing. The speed of a single job will never be faster with multiple THORs, but the throughput can be. You can run two THORs picking up jobs from two different queues or the same queue. Its a configuration. The downside of running multiple THORs in the same hardware is that the physical memory on the nodes, needs to be shared out amongst the THORs and this needs to be configured per ThorCluster definition. You wouldn't want multiple THORs on a h/w which didn't have enough CPU cores to support it either. Not recommended ever having >Thor's than # cores.. Probably max should be something like #cores / 2.
NEED MORE INFO ON HOW MULTIPLE THOR CLUSTERS ARE BROUGHT UP IN THE SAME HARDWARE.