MIDI Problem with FL Studio and energyXT

[Spirograph]

I've been trying to get all my hardware to work with my software but am running into some problems within FL Studio.

First, my setup:
- Evolution UC-16
- Evolution UC-33e
- Evolution MK-361

All of these are run through MIDI-Ox and go through MIDI Yoke 1.

I've configured energyXT to support all my MIDI hardware so I can control my VSTi's. The basic setup is that each energyXT file holds one VSTi that has MIDI channel filters on it so it only receives MIDI data from the specific channel that corresponds with my hardware. Additionally, each VSTi will receive MIDI data from channel 16 which is what my MIDI keyboard is set to. For instance, Claw receives signal from MIDI channels 6 and 16. Channel 6 has the correct CC settings so that recalling preset 11 on the Evolution UC-16 gives me full control over Claw. This works fine when energyXT is used as a standalone and has worked well in Tracktion as well. However, when I load energyXT up in FL Studio, I run into problems. I can only get my keyboard to trigger notes if the channel is set to 1. This also applies to sound, so if I have a MIDI score loaded for the channel, I can't hear anything unless the channel is set to 1. Also, none of my controllers seem to work for changing the settings on the VSTis loaded within energyXT.

The reason I use energyXT is so that I can have one program control all my VSTis and then be able to load energyXT in the host of my choosing. I dabbled with Tracktion for awhile but ultimately decided to wait until v. 2.0 til I consider it seriously. I was hoping to incorporate energyXT within FL Studio since I find it a pain to have to constantly reconfigure my VSTis with my hardware.

Does anyone have any suggestions to what I'm doing wrong? I would also settle for a solution involving being able to configure a VSTi in and have those settings work everytime I load it up in FL Studio since it is my main host of choice. I tried a huge template file with all the VSTis I have fully configured, but it proved to be too CPU heavy (no real surprises there).

Any help is appreciated.

S-graph