This has been an ongoing issue since last fall. I am out of ideas and everyone I talk to about it is stumped so I am looking for other ideas here. I will try to list everything that might be helpful but feel free to ask questions. What ever it is, it is over my comprehension at this moment.

Engine:
LS3, stock bore/stroke
ported heads
Comp cams 54-453-11

The EFI system uses a separate O2 sensor in the passenger header for feedback, then I have one more sensor in each header for the gauges.

The last time it ran good I was attempting a burn out but the car bogged down. Drove it home with what felt like a dead cylinder. Ohmed out the plug wires and found an open one, changed it out and checked the plug, it still looked decent. Ran it in the driveway and it seemed good. Next day I drove the car, got a couple blocks away and had the same symptoms. Over the next few weeks I changed all the plugs, swapped coil packs and fuel injectors from side to side, ohmed out the coil pack harnesses, removed intake (checked for a bad o-ring seal), checked for vacuum leaks, replaced all vacuum lines, swapped O2 sensors side to side, relocated O2 sensors into the collector (previously mounted a few inches downstream from the collector). The issue was still present on the same side of the engine. Compression test and leak down tests were all decent.

Here is a video of what was happening.
First 12 seconds was cruising, constant throttle, 13-17 seconds light acceleration, 18-23 decel then back to a cruise.

Had an Atomic MSD efi system in the car, at that point I thought it was possibly a faulty fuel injector driver circuit. Called them up, told them what I have done so far, they agreed it was possible. Sent it to them, they tested the system and it was found to have no issues. In disbelief I bought a different system, purchased a FAST EFI 2.0 and installed it. Exact same issues. Started thinking it might be a mechanical issue.

Removed valve springs and had them checked, thinking possibly one is week and I have a valve not closing properly. Springs were all good. Removed the heads and valve stem seals, found I had some very sloppy valve guides. The engine was rebuilt and has about 3500 kms on it. Almost all of the valves had noticeable movement like this.

Worn valve guides:

While the heads were off the cylinders looked good and the tops of the pistons had no damage from valves or detonation.

Heads had all new guides put in, noticed 2 valves were bent while lapping so they were replaced. Removed the large roller rockers in favor of an OEM style with a trunion upgrade so that I can run the OEM valve covers with baffles for the PCV so the catch cans don't get so much oil in them. Installed with new pushrods and valve stem seals.

This video is from getting it back together yesterday:

This is from start up this morning at room temperature:

One thing I have noticed every time is that it doesn't seem to be an issue until the engine has some heat in it, around 100*F is when it starts to go lean.

 

No exhaust leak letting air in?

Ls banks never read the same

 

I pulled the headers off and checked for cracks and visible damage, couldn't see any. Also put on new gaskets when I installed the heads, there were no carbon leaks on the old gaskets.

 

This is my dyno sheet, banks are always different

 

Exactly rite is the reason I use only one wide band after the x pipe where all cylinder mixture have joint to get an average overall reading and tune with it.

 

I'd start at the very baseline, the engine and the dyno. I'd like to know on what planet a peanut cam stock stroke ls3 with stock heads makes 700+ engine hp, because that isn't happening on this earth. And peaking at 8k? lol. someone's rpm pickup is off...

 

Thats a good idea, but thats not my dyno sheet. I haven't dyno'd this engine yet mainly because of the large distance to the closest dyno around here.

 

is the driver side leaner than the passenger side is what im understanding? if so, I had read in the past in other threads that the LS motors generally run leaner on the driver side for reasons I can't remember.

 

If I have the O2 sensor for the EFI in the passenger bank, the drivers side will be leaner than 15:1 when accelerating and have a loss of power and lots of engine vibrations like dead cylinders. If I put the EFI sensor in the drivers side then the passenger side is pig rich. I wouldn't be concerned if it was just 0.5-1 difference and still ran good.

Going to take it to a shop once the snow melts and use a different system to fill the intake system with smoke, see if there is a leak.

 

so you're saying when you swapped the EFI sensor from passenger to driver side the leaning issue (loss of power and engine vibration feeling) goes away?

 

^if so, have you thought about replacing the EFI sensor? could be a faulty sensor. in the last vid you posted "cold start-up until warm 100*F* was the EFI sensor the one on the left?