The desired magnetizing current is up to you. You can reduce it but then the maximum power the transformer can deliver is also reduced. I think the function is roughly reciprocal ie. Imag = f(1/N) - try plotting magnetizing current as a function of the number of turns for your expected primary voltage.
I remember reading an article somewhere that described the rationale behind MOT design. I thought it was this one:
https://www.qsl.net/kh6grt/page4/xfmr/xfmr.htm but it's not. It's worth a read if you haven't seen it.
I understand that for the most part, MOTs run at full secondary current except during the time that the magnetron's heater is warming up and the anode current is reduced. Because the secondary load current cancels out the magnetic flux developed by the primary current, increasing the secondary current moves the transformer away from saturation. The designer can then use a smaller core that can saturate at a lower primary voltage than for equipment that does not operate at full current.
Bottom line, if the transformer will not operate at high load current almost all the time, the losses due to the magnetizing current can be significant.