I looked at the specs for both the Zephyr Star and the Luxman 590. There is no obvious reason why this pairing cannot be made to work well. First, I don’t understand quite why SoundSmith recommends a 470 ohm load for this cartridge. It’s internal DC resistance is only 10 ohms, so in theory it should be able to drive the 100-ohm load of the Luxman MC input stage. The problem might be that this is not actually an MC cartridge; therefore it has an inductance much higher than that of a typical MC. (About 100X higher, if a typical LOMC cartridge has an inductance of about 25uH. Inductance of Zephyr is 2.75mH.) Maybe Atma can comment on how that would be an impediment, but perhaps that is also why SS recommends a 470 ohm load. We also know from specs that the output of the Zephyr (0.4mV) is not too low for the MC inputs to handle, because Luxman tells us that the MC gain (NOT stated on the Luxman website but can be calculated) is adequate for a cartridge with an 0.3mV output. Before doing the research, I had assumed that the MC gain of the Luxman is derived from using a SUT placed ahead of the MM stage, but the website, while very lacking in phono-related info, suggests that MC gain is done by an active stage. For one thing, in their many photos, I see nothing on the PCBs that looks like a SUT. OK, so if there is no SUT, then there is no problem to replace the 100 ohm load resistor with a 470 ohm load resistor. Any skilled technician can do that. I would do that way before throwing either baby out with the bathwater. I am a big fan of MI or "fixed coil" cartridges.
Oh, by the way, this problem cannot be fixed by paralleling an external resistor with the 100-ohm input resistor, net parallel resistance is always lower than the value of any one resistor in the bundle. So paralleling an external resistor can only lower the net input resistance, not increase it.