Hi,
my suggestion is start at the cutting tool and work backwards.
By this I mean you need to determine what materials you want to cut and what tools you want to do it.
That will in turn define the power, torque and speed of the spindle. Then you need to decide on the maximum cutting rate
and possibly also on the traverse rate.
Those decisions will in turn determine the required rigidity of the axes and therefore the technology you use in them,
be it round rails and linear bearings or rectangular rails and recirculating ball bearing loaded cars or some other method.
The same decisions will determine what size steppers are required, or if the speed requires servos rather than steppers.
The cutting forces and acceleration/deceleration forces will determine the torque required of the stepper/servo.
You will have to go through the decision tree several times because you might decide for instance that you want to cut steel
but as you progress through the decision tree you realize that the cost has just blown out and its un-achieveable.
I need to point out that I did not really do this when I designed and built my mill....but I rather wish I had. I tended to
settle on those components that I could get reasonably priced second hand. Its very easy under those circumstances
to find an 'overkill' part or assembly there but only a sub-standard part there. Trying to keep the design balanced and
prevent the budget from blowing out making decisions from the tool backwards in invaluable.
Craig