When faced with a pile like this, I resort to Linux for anything I can hook up to something that will run it (SATA, IDE en SCSI no problem, MFM, RLL and EDI may be more of a challenge as halfway modern Linux systems need 686 at least)
Under Linux you don't need to partition or format anything, you can just run a low-level check with the following command:
badblocks -nsv /dev/[name_of_hdd]
You can even do multiple simultaneous checks, although I'd recommend not doing more than one drive per IDE or ST506 channel simultaneously (one check fully saturates the interface, so two at same time will not be any faster). Depending on how many drives you can do in parallel this speeds things up by a factor two to four or more compared to testing under DOS/Windows. If you don't have a Linux box with IDE, SCSI and SATA available (I tend to use either a P3 with SATA and SCSI cards, or an early Athlon64 with SCSI card), get an old Knoppix CD image (from 4.x or 5.x release) and boot from that. It's a liveCD distro, so just boot from CD and a few minutes later you have a working Linux environment.
With MFM/RLL/ESDI drives the challenge is that it's non-trivial to get the controller for that stuff to co-exist with an IDE controller. What I'd do if I had a big pile of disks like that to test, is to get a 486 motherboard without onboard IDE, add a SCSI card with its own bootROM for the OS, then install an old Debian version (2.2 Potato for example) on that. Once that's working, add the ISA MFM/RLL/ESDI controller, hook up the drive, get it configured in BIOS and see if it can be addressed from in Linux (most controllers will be supported out of the box).