Mostly I stay out of "MOSFET-mod" arguments.
Nobody wants to hear my opinion, and that's all good.
One thing I observe that's not an opinion is a particular controlling factor in how any MOSFET behaves, no matter where it's used.
The DC gate-bias voltage on conversion-job MOSFET mods is a bit of a wild card. Too much and you get really high wattmeter readings, followed by one of two things.
1) Blown fuses, or
2) Plastic face blown off the MOSFET when it overheats and fails.
Too little DC-bias voltage on the MOSFET gate and it's just weak.
Factory-installed parts and circuits found in a radio built with MOSFET finals is the result of an engineer's tweaking, to get a consistent result from thousands of units. If you make a slavishly-indentical perfect copy of that circuit in a conversion radio, it should perform the same.
A one-off conversion might fall neatly into place and function just great. Or, it can fall victim to factors nobody can predict.
I have a suspicion that radios frequently get this mod because the power was below par in the first place. If that problem was caused by reduced drive upstream from the original final transistor, the new MOSFET will also be working with reduced drive. Too many possible causes of that to make any simple rule.
Except one.
Only convert a radio that's delivering normal, healthy transmit power.
A radio that's already weak, compared to normal power for that model, may still be weak with the conversion in it.
But if it's giving you the normal, proper power with the stock transistor, where is the incentive to modify it?
But yeah. The radio most likely to give you good power is a radio that doesn't really need the conversion.
And a radio that gets converted because it only did half or 2/3 the wattage it should may have gremlins in it that hold back the new MOSFET final as well.
Toss of the dice.
73