Hello again,
If i understand you correctly, that sounds right. Lets assume that the primary is now completely open circuited so only the secondary is connected. It is entirely possible that the rest of the circuit connected to the secondary can generate AC voltages high enough to saturate the core and thus lower the impedance of the secondary winding. But what you would do here is you would analyze for that operation point. If the secondary can take say 100 volts and you only see 50 volts with the normal secondary resistance and inductance, then it is not saturated. However if you see 200 volts then you might be concerned about saturation.
So you would first analyze it assuming everything is ok before the analysis and then actually check to see if is ok after you see some analysis results. If it is not ok you'll have to look into it further.
You might also note that superposition assumes a linear network, and saturation is not considered linear. However, if you find that it is not in saturation by the first analysis, then you might assume it is linear.
So when you think about it, this is almost the same as any other analysis. If you have a simple single standard 1/4 watt resistor being powered by a current source and you analyze the circuit and find that there is 1000 volts across it, you'll have to modify something somewhere because that little 1/4 watt resistor can not handle a voltage that high (normally they are rated for 200 or 300 volts no matter what the resistance is). This is even more immediately apparent with a capacitor. If you have a 1uf, 25v capacitor in a circuit and analyze the voltage across the capacitor and it comes out to 50 volts, you know it's not going to work. You either have to raise the voltage rating of the cap or modify the circuit in some way.