If the arm is moving at all, which it is, then the brake is disengaged. I'm pretty confident that the brake is no longer a factor here.
My advice: be patient. In the same way that you've exercised the spindle so it moves freely again, it's time for the arm to wake up in a similar way. Rather than move the arm yourself (like how you bump-started the spindle), what I think you should do is repeatedly reset the drive so that it keeps attempting to move the arm all on its own.
My bet is that as you do this, you'll find that the arm moves a little bit more with each reset. If this works, it could take a while before the arm regains its full range of motion, but in time it might get there. This is something I've seen happen before.
To reset the Widget, you can always reboot the computer (if I remember correctly, I'm not 100% sure): the reset button at the back of the machine is the simplest way to do that. It may also be possible in NeoWidEx via the
SOFT RESET command, but I'm not so sure.
If the arm does start to free itself up in this way, things may develop like this:
1. The Widget attempts its "post-reset arm wiggle" a few times and fails.
2. Finally the Widget gets a successful wiggle and decides to start the surface scan. But it can't make it through the full scan and eventually fails.
3. With successive resets, the scan can go further and further.
4. Eventually you get a complete scan. The arm probably could still do for more exercise, and for that some more resets is probably the easiest way. You can also use the NeoWidEx
SCAN command to repeat surface scans, which of course move the heads all across the disk.
Once the Widget seems to be revived, as I hope it will do, I recommend getting a serial cable and a copy of
BLU and dumping the Widget's contents before you attempt to boot from it.