Here are some thoughts...
In terms of outboard passives, there is at least one advantage. If you ever consider going active, you are all set. Just unplug the passives, plug in the active amps and you are good to go. Otherwise you would have to take the speakers apart, disconnect the crossover from both the binding posts and the drivers, and rewire directly from the binding posts to the drivers. You cannot leave the passive crossover in place when you go active.
There are some who would say removing the crossover from the speaker prevents internal cabinet vibrations from impacting the performance of the crossover. I haven't seen any objective data to support this point of view and have my doubts. While I can't say for certain that this would have no impact on sound quality, it would certainly not be dramatic. (I've heard quite a few HT3's done both ways and have not noticed a difference.)
In terms of bi-amping, the amps must have the same gain structure in order for this to work without going active.
But the question I would ask is, "what are we trying to accomplish here?" Keep in mind that the in a 2-way, the tweeter will draw relatively little power compared to the woofer. So freeing the woofer amp of any duty supplying tweeter power will not likely result in much improvement. In a 3-way design that plays much deeper, it makes more sense to drive the woofer separately.
I you want to run tubes on one section and solid state for the woofer, that is another question. I have always wanted to try tubes on the midrange and tweeter of a pair of HT3's with a large solid state amp driving the woofer. But that would require a tube amp and SS amp with the same gain. (I have talked to Frank Van Alstine about modding an Ultra 550 to bring its gain down to the level of his Ultravalve - that would be a very interesting experiment).
In the end, however, if you want to use separate amps for each section, the best way is to go active. Then you do not need to worry about gain relationships or driver sensitivities - the processor will handle the job. This, however, is a much more expensive proposition with the cost of the processor and extra amps added to the mix.
Every once and a while someone asks me which direction I think they should go - active or passive. From my perspective, with the quality of Dennis Murphy's crossover designs, you should not expect significant performance improvements from the midrange on up - there just isn't much to improve upon as the passive crossovers leave little room for improvement.
In the bass region, going active will allow you to accurately apply EQ with room modes, but the cost for this type of setup is high.
The only other advantage of going active is being able to use one type of amp for for the midrange and/or tweeter and a larger SS amp to drive the woofer. This approach may have some merit. But, again, the cost will be high, relatively speaking (we're talking here about perhaps doubling the system cost for what may be relatively little gain in performance).
In the end, it comes down to the level of performance you wish to achieve and the amount of money you are willing to invest. I would determine the latter first and then develop a strategy to get the most for that investment. As an example, the cost for a totally active HT3 setup will be in the same ballpark as a passive HT4. It wouldn't take me long to make that decision.
- Jim