I have a few of these motors, and a project I have built to use it. I just have to wait for the place to set it up.
The stator of that motor is completely sealed, almost impossible to get to the connections to try anything different than star / parallel coil original arrangement. I disassembled a 05A2AB and tried to remove the plastic / epoxy stator casting on the rear to expose the junctions. I did find them, but damaged coil connections at the same time. It wonh't melt or burn off, I could only heat and chisel or grind it off. SO don't bother.
The shaft drag may be the front shaft seal, and the bearings are very stiff due to the type of grease inside. I used the flanges from the destroyed 05A2AB and flushed the bearings and replaced it with a lighter oil. Used them for shaft bearings on a tower. I don't think you need to do it, the bearings will probably never die if they are not tampered with. And that front shaft bearing is a monster, rated for high radial and axial loading (i.e. throw a big prop on that thing). The bearing pictured is the rear shaft bearing, but note that large shaft diameter on the front, the bearing that fits that is 32mm ID (1.25") and over 3 inches OD.
Below is a picture of a disassembled 05A2AB (same series as yours / no brake / 450 watt size) and an open 09DA1S (same specs as yours / older series / double the size). I also have one identical to yours without the brake. The rotor of the 450 has two offset rows of 8 magnets. I never pulled the rotor from the older 850 or the newer one.
I did make a simple bench test of several of these motors using a variable drill, a clicker and stopwatch, and a fluke meter. Not scientific but at least gave me a basis to compare the motors. From my test your motor will reach charging voltage (for 12V app) at around 200RPM. The results may be current-limited because I used load resistors with 50W rating, and the upper speeds exceeded that wattage, producing a thermal generation at the testing component that could be registered on touch-sensor (i.e. it burned my finger). This means upper measurements may be lower than actual possible output at the given RPM (I hope).
And a picture of the new 450 and 850 with specs for them and additional 450 and 750 small / 3000RPM motors:
What I didn't include in the tests were results of re-wiring the older 850 motor (large frame) to a star / series coil configuration. As expected, it tripled the output voltage. Gonna use that one on a VAWT with a 1:1.7 ratio chain drive.
If you have a choice between an older motor and a newer one of the same wattage, and the size difference is not an issue, the older one may be more adaptable - depending on your application I guess.