Wideband AFR AutoTuner. No Need For Tuners?

I found a cheaper and possibly more effective way for an autotuning setup.
I have an Innovate MTX-l reading wideband. It has a programmable output for ecms, third party data loggers etc. If you set full rich via AFR to .96v=12.8afr
And full lean to .00v=15.0afr in the output tab in LMProgrammer, this will tell your ecm to aim for 12.8 when you full throttle or other times the ecm wants to be full rich and the ecm will try to avoid anything close to 15.00 unless when cutting throttle
I then went ahead and connected the MTX-L output to the grom's O2 sensor input in the wiring harness. On the right side of the bike next to the starter relay
This will give the ecm a better read out on your setup than the narrow band sensor it has stock.
Ive ran it like this for over 1000 miles so far, no check engine and the bike hauls ass. I can literally see the ecm adjustments on the gauge. Before when i had the harnesses O2 input shorted like how the bazzaz zfi recommended, i had to manually adjust for being at full lean. And the ecm sometimes would make random adjustments, presumably from the ground possibly getting interference.
I have not tried running it with a cleared map to see if the tuner would even be necessary. Ofcourse down the line with the more mods you do, if you dont upgrade the injector along side, it would lean out if you max out the ecm increments say with a 4v or so id believe. I dont know anyone else that has done this but the bike is alot more stable with afr readings, no random lean idle, random rich bog at full throttle. The ecm is learning a new map i flashed 2 days ago so its got its hiccups, but for the previous 1000 miles i didn't have any. Wiring for the tuner and gauges is critical. You HAVE to have good connections. Consistent voltage, no excuse. It will vary the voltage readouts the ecm, tuner, and wideband gauge will pick up.
My theory, is if someone does this with an upgraded injector, a tuner isnt even necessary. The ecm would have a wideband input to go off and can adjust accordingly regardless of the setup you run.
If someone can try that theory, it would open HUGE doors and possibly start eliminating the need for a tuner overall

So I'm having a super weird symptom with my Grom but being that the Grom part of my bike is practically gone and this now has become my own Frankenstein I'm not sure who's going to be able to help me with it. Worth a shot tho.

I have a bazzaz zfi Unit,
TB Cam,
Flipped TB To face forward,
Stock head, piston, cyl, crank
Amazon exhaust,
no intake pipe. Just a pod filter to the TB

So a little while back I put my innovate mtx-l to talk to my ECM via programmable output to the ECM O2 pickup. The thing practically auto-tunes now. Runs like a champ with it. The issue is it's consistency. Personally I feel it's a bad connection somewhere the main problem is finding out where.
Essentially what the bike is doing is leaning out randomly on the narrow band but Wideband seems to freeze for a sec or two and the bike WILL lean out while the wideband is still reading normal afrs, then it'll return to normal running. The bike will feel a little off after return and almost feels a little down on power but I've had flat tires and head wind the last rides so it could've also been that. When the bike has its episode, the Bike will sputter like its rich but the narrowband shoots to L (+17.0AFR).
Questionable connections

CKSPS to BAZZAZ UNIT
TPS to BAZZAZ UNIT
Neutral to BAZZAZ UNIT

The 12v power source to the bazzaz unit MAY or MAY NOT flicker due to a bad connector but I'm not sure how that would cause the bike to shoot lean out of nowhere, hold it anywhere from 1-5 seconds and then return to a slightly less regular state. Ive disconnected the cksps while riding and it WILL shoot to lean in closed loop but it also reflects on the wideband. Narrowband as well.
When it has its episode and shoots lean, the narrowband is the only gauge to pick it up.
One thing I've noticed was once on a ride, it felt like it did the opposite. It felt boggy and when I looked at the wideband, it went from 12.8-13.0 AFR (regular) straight to 10AFR with no change to how the bike was running. I didnt look at the narrow band to see if it was rich but something is up with the wideband and the bike that is new to me and I'm not sure where this originated from. The O2 does have some miles on it and it wasn't completely excluded that it could be a failing o2 but I haven't had this kind of behavior before and usually the gauge will read an error when the O2 goes. This is makin me scratch my head, not gonna lie.
Open to what ever ideas you got

UPDATE: I've tested the mentioned connections and I don't see any inconsistent values in the bazzaz software. All values do respond accordingly and tugging and fiddling with all the wires did nothing. I don't think they are bad. I swapped the o2 for another one i had on the side (used) and it did the same thing even still. I'd ride around in second to third gear at around maybe 40% to 50% throttle and wideband and narrowband both read regular values (12.8-13AFR Wideband, 12.7-13.AFR narrowband) Then suddenly it'll sputter and the narrowband shoots lean while wideband stays 12.8. i don't think its an o2 sensor but being i haven't tried a new one yet it can't be ruled out. I'm gonna run thru the o2 wires going up to the controller today to see if there may be any cuts or faulty wires that may cause a false reading on the o2 side cause false adjustments to the ECM. I think the wideband is sending false info to the ecm and the ecm is leaning out. I'm not sure why it would be intermittent tho and that's kinda whats making it hard to track. I'll update with what I find if anything