I do use subs, for the reason being that what and how they do it isn't found in integrated form in a pair of floor-standing main speakers (the subs take up 20 cu. ft. per cab, as a FIY). The main speakers are large (6 ft. 2" in height) as well, and in conjunction with each of their corresponding sub is considered and treated like a single speaker system per channel. The system covers down to 20-25Hz at full tilt.
If that's an acceptable outset for the criteria set by the OP (again, essentially they're large floor standing speakers divided into two cabs per channel, treated and configured as a single entity), here's the deal: they're "EQ'd" actively over the entire frequency range by the same, quality DSP unit, and as such - combined with placement and acoustic measures - gives you an elaborate way to "morph" the performance at the LP in all aspects of filter settings - on the fly. Normally one isn't granted such a plethora of EQ'ing possibilities, and thus the performance of speakers is mostly rooted in their pre-configured passive filters (and room integration) with any EQ'ing done "on top" of this as an extra measure/layer, either via DSP or some form of analogue equalizer device.
I find having the opportunity of described active EQ'ing capabilities to be indispensable. We near-field measured my main speakers for notch placements, and measurements at the LP indicated PEQ areas in the filter settings. Everything from hereon was done by ear; fine tuning Q's, gain structure, delays, crossover points, slopes, overlapping/not overlapping (asymmetrical) XO points, etc.
I've tried DRC FIR-filter in both the time and amplitude domain over my previous, passively configured speakers, but wasn't entirely satisfied with the results as I found there to be a noticeable processing imprinting. As is I prefer the more "manual" by-ear approach aided by measurements, actively - meaning the EQ'ing is applied at the heart of the speaker's intrinsically necessary DSP crossover as the only signal layer.