I hadn’t looked at the BMS data sheet in a long time. It’s a 118dB/2.83v driver. Most of the people I work with are using Klipschorns and LaScalas, which have higher sensitivity. So a bit more room to work with. The issue relates to balancing the drivers.
Klipsch will show the Cornwall with a K-33 woofer as 102dB/2.83v. There is a footnote on the spec - “in average listening room”. What does that mean? It means Klipsch adds 4dB of room gain to the specification. The assumption is that the owner will place them based on the set up instructions. To get that, you need boundary reinforcement, i.e., walls, and preferably corners. Most people however, normally have them on a single wall, pulled out, and turned in to improve imaging. They lose half or more of this boundary effect, or gain - which is typically in the bass.
I noticed a comment related to trying a different woofer. I would be really careful here. There is a mathematical alignment between the driver, the cabinet volume and the ports. Unless you have a way to measure what you are doing, I would stick with the recipe that is tried in true. Use the K-33 or a close equivalent.
The old spec for the Cornwall is 98/99dB. That doesn’t sound like much of a difference, but it is. I split the difference and call it an even 100db. So, the midrange and tweeter need to be pulled down to match this. This would be voicing for flat(ish) response. However, many prefer a tilted down response, and this relates to the equal-loudness curves.
The horn on the driver problem: when you look at a data sheet, you will see a frequency response. This response was taken on a horn of the manufacturer’s choosing. When you change the horn, everything changes. The response, impedance curve, and sensitivity - or gain. You simply do not know what you actually have unless you measure. It is usually pretty close to what you figured it would be, but it is not perfect. The Universal is a good network, but it is not what I would call “optimized”.
With that said, it’s important to stay within the bounds of things that people have done before that have been measured and we know works. The more you move away from that, the less you know about what you actually have - unless you measure.
The BMS problem: the 4592-mid is a great driver, but it is 118dB/2.83v. To bring it down to 100dB to match the woofer, you need to knock off 18dB from the midrange. Having an autoformer that takes off 18dB in 1dB increments is not really helpful here. You have bottomed out and have no wiggle room for further adjustments if you think the midrange is too forward.
The B&C DCM50 is a better choice for this application. it uses a composite diaphragm that everyone seems to like the sound of. I have never read a negative review about this driver. At 108dB/2.83v, it leaves quite a bit of adjustability in place for dialing things in.