If you use a developer, you typically change its pH, use up some of its development agents and build up some bromide. A replenisher can bring pH back up to where it should be, it can add the missing amount of development agent, but I am not sure how you would deal with the extra bromide. If you look at XTol, it uses Phenidone instead of Metol and is therefore less sensitive to bromide buildup, and still the replenished XTol setup is said to yield different results from fresh stock. You can compensate the bromide buildup by raising pH a bit more, by adding more development agents than what is actually lost, or by adding a strong solvent like DTOD in small amounts just to compensate for the bromide. Oh, and it's not at all clear from the onset, whether you need to replenish more Metol or more o-PPD, and whether you can find an affordable source for o-PPD if you can't make that replenisher off W665 concentrate/powder.
My initial approach would be to replace/add about 100ml developer per roll, then measure and restore original developer pH. Use original dev time and check, whether your developer is still too inactive after this. The results should be somewhere in the ball park, so don't do this for critical stuff, but you should get somewhat useful negs anyway, so no reason for endless test strips. If you think, that your replenished developer is too slow, add tiny amounts (increments of 100mg/l per roll) of DTOD to the replenished developer until its activity matched the original thing.
If the results suit you, then you have something like a replenished W665. If the results don't suit you, then Fotohuis may have been correct in stating "don't replenish W665". You'd have to experiment ...