Closed loop microstepping - similar torque penalties to open loop? Questions...

If I buy an 12nm closed loop stepper, like a leadshine closed loop clone, can I use 1/2 microstepping and not go down to 71% torque, or 38% for 1/4 step?
If I buy a closed loop stepper, can I have less of a 'reserve torque' and thus probably go down to a smaller stepper motor? Perhaps a 8Nm where 12nm are commonly used (e.g. an 8Nm motor for the z-axis on an rf-45 clone)?

My plan / reasoning / ideas:
I plan on using leadscrews for my RF-45 clone of 5mm, which translates to 0.025mm movements for 1 a step on a 1.8 degree stepper. Seems too large, I want to hold 0.02mm or less, so I need to microstep right? I am thinking 1/2 microstepping would be ok as it would bring me to 0.0125mm per 'step'.
I do this by setting the DIP switches on the driver, which sets the microstepping to some fixed value I guess.

Is my logic right? Maybe a 10mm lead with even more microstepping would be good?
Incidentally anyone know if the leadshine clones are good, or should I pay an extra $100 per axis for genuine leadshine.

Thank you!

I used 5mm pitch ball screws in my setup. I have some closed loop steppers, and they are no different in performance than their open loop counterparts. The closed loop stepper and driver I used can be ran a little 'looser'. In other words, I could run it faster than the open loop version and if it missed the occasional step it would pick it back up. But if it missed more than one step, it would fault and stop. There are different configurations for different closed loop drivers and I didn't experiment with many of them. Some will actively resist moving and if you do get it to move it will move its self back to where it was. I found out rather quickly that during a feed motion if a stepper started missing steps or move back a step that my part was pretty much toast. Making up the steps later was too late, because the other axis may not have missed/gained the same steps. you end up with a flat place on an arc, or an egg shaped hole. In the end I decided it wasn't worth the extra money to use closed loop steppers. I would have just went with servos.

The open loops I run more conservatively so that the chances of it missing a step are as close to zero as I can run it. The faster they run the more likely they are to lose steps. Make sure you have an adequate power supply for your motors. The higher the micro stepping the more likely it becomes it can lose a step as well. With a 5mm screw and 800 steps per revolution (micro stepping on the driver, the motors are 200 steps per revolution), I get 160 steps per mm which gives me a theoretical resolution of .00625mm. I could get more speed out of a 10mm screw, but less resolution.

I bought a leadshine driver and motor, and I also bought 'clones' The clones seem to be just as good and I used 'clones' that are allegedly made by leadshine and I have had zero issues with any of them. If i had any complaints about any of the steppers I have used over the years, it would be that the bearings arent all that great.

I am running 480 ozin NEMA 24 steppers, which are substantially smaller than 12nm motors. I lift a 108 pound z axis up without any trouble. Im not sure how large of a machine you plan to build, but you get a significant mechanical advantage with the smaller pitch screw, and Im not sure you are going to need a motor that large. A large motor requires more power and a bigger driver to realize its potential. A poorly driven/under powered nema 34 wont run as well as a tuned nema 23 or 24 setup. I even have some nema 42 motors, but they require a massive amount of power to run correctly and they perform very poorly compared to my 'little' nema 24's because I just dont have a power supply large enough. Just keep in mind that the motor isn't working alone. The screw is in there and its giving a mechanical advantage.