It will never be as simple as "I specify input and load reactors on all drives".
Let's just think this thru for a minute. If you have dual voltage motors and are running them on the lower voltage, 230V or 190V, load side reactors give you absolutely nothing in terms of protection and actually waste some of the system energy as heat. To say nothing of taking up space and costing real money.
Next, if you have spent your precious cash for Insulation Class F motors with MG1 Part 31 endorsement, you are good for 60 feet motor lead length at 10hp and 250 feet motor lead length at 100hp. You can extrapolate between those points linearly. Why spend the money for a good motor and then spend it again for a reactor? Further, motor lead reactors make it much harder to do speed and torque control in sensorless vector mode so they are to be avoided as much as possible for those applications.
Further, if you really have longgggggggggg motor leads or the potential for common mode noise, a reactor may make you feel all warm and fuzzy but, in the end, it will still keep you up nights with false faulting, ground noise and motor insulation damage. Unfortunately, in this field, you actually have to think, analyze, and then respond appropriately sometimes.
As for input reactors, first and foremost, reactors drop voltage at rated current, 3% of the voltage for a 3% reactor and 5% for a 5% unit. If you are in a low voltage invironment to start with and you automatically add a reactor just to "feel good", you just made your low voltage problems worse. And, how about all those specs written "5% reactor required ahead of ALL drives" when the drive is already built with 3% reactance internally. That's 8% voltage drop at full load! Let's see here. 460V times .92 is 423V. Maybe this isn't so smart after all! Did I mention that these things cost money and waste energy.
Further, let's say your drive system is to be installed on one of these miserable floating delta power sources. You can add all the reactors you want and you will just make the common mode noise problems worse. You need a drive isolation transformer with a grounded wye secondary to give the ground pulses a short path back to the DC bus. Reactors don't do that.
Further, if there are PF caps upstream somewhere (no, the customer will not tell you where) and you add reactors, you risk putting the power supply into resonance. Probably want to avoid that if you don't really need reactors to start with.
Some manufacturers are now starting to add swinging (saturable) reactor in the DC bus instead of input reactors. These will give you the harmonic suppression you need just like an input reactor but without the high load voltage losses. Very crafty! What they don't tell you is that the swinging choke doesn't protect your input against voltage spikes nearly as good so that could be a negative.
I guess I've said enough on this. Bottom line: you have to think about what you are doing and maybe even analyse or take data. The alternative is to waste money and energy and simply kid yourself about the results.