Do you prefer more woofer or more CD/tweeter at xo?

When you’re fine tuning your crossover for that last bit of voicing, do you prefer more woofer or tweeter at the crossover point? Assuming you’re dispersion is good either way, and response isn’t negatively affected. Say add 1db to one side and take away 1db to the other. Which driver is typically the better option to use more of?
 
Maybe I can describe my question a little better. When you have a crossover combining the output of two drivers, they both contribute to the total output over a range. You can get a flat result while using more or less of either driver. Turn one up and the other down and still get the same result. When you change the level the off axis will also be affected, but not a lot if we're talking a 1 or 2 db. So the crossover point can stay within a 100hz region but have more or less woofer or tweeter. I've been bouncing back and forth between the two and can't decide which I like better. The difference is subtle, but it is different. You should easily be able to duplicate this in xsim by simply adjusting the component values of each driver to match in two different ways. I've been able to achieve the same final response while changing multiple different components on each driver.

Hopefully that explains it better. Wasn't sure exactly how to say it.
 
All depends on the drivers. A lot of people push their tweeters too low causing undue distortion, but on the other hand, mid-woofers tend to get ragged before roll-off up top. 6 inch a problem, A 5 will do better, but that leaves them lacking below unless we use several. A 5 seems like overkill to cross to an 8, but maybe.

Some woofers, like aluminum, CF, ceramic, have horrendous breakup mandating a steep cross as low as possible, so letting it "fill" for a phase miss-match may not be wise. Then all the people, some well respected names, try to use a 1 inch dome with their natural roll off, 1500 or 1700. Pretty FR chart. Bad sound.

So the only correct answer is "it depends"

I am on the verge of going with Adson and trying to find a decent mid. A 6 is barely enough for the mid base and it can't quite reach a 1 inch dome. Most tweeters really skyrocket distortion somewhere around 2K. Just not found that magical mid with a smooth response, low distortion and enough efficiency we don't have to pad the woofer. The compromise may be a D'Apolitto aligned MTM, but he requires a 4th order crossover which I would prefer to avoid. Getting phase right in a 2nd is hard and a hybrid 2nd/3rd has it's issues too but can get close phase on a flat baffle. I have not explored what a recessed/waveguide tweeter and 2nd order MTM would work out as using a pair of 5's that can make it to 3K. I run a sub.
 
Maybe I can describe my question a little better. When you have a crossover combining the output of two drivers, they both contribute to the total output over a range. You can get a flat result while using more or less of either driver. Turn one up and the other down and still get the same result. When you change the level the off axis will also be affected, but not a lot if we're talking a 1 or 2 db. So the crossover point can stay within a 100hz region but have more or less woofer or tweeter. I've been bouncing back and forth between the two and can't decide which I like better. The difference is subtle, but it is different. You should easily be able to duplicate this in xsim by simply adjusting the component values of each driver to match in two different ways. I've been able to achieve the same final response while changing multiple different components on each driver.

Hopefully that explains it better. Wasn't sure exactly how to say it.
If you're changing the amplitude of either driver, not just in the transition, then you should consider how it will effect the whole response. Might sound too bright, lacking bass or vice/versa. Ultimately, how it sounds to you is what counts. I'd try a variety of music tracks to help decide. What is your driver compliment?
 
In this case it is more about the horn and how the driver sounds to you.

Looking at the impedance curve when loaded. It can go down to 1.6k
But is likely to have far less distortion @ 3.3k

technically you would like to cross over much much lower.
but the listening test is more the judgment.

1.6K, 3.3k and 5k have always been rather typical points for good reasons.

as many have mentioned if there is fatigue or frustration with distortion.
Some have moved towards 3 way systems. and you can get that driver far far away
from impedance peaks @ 5K