I'm going to have to check again but I have an In car vacuum gauage and I'm pretty sure that was reading the same . If the carb signal had been effected I would think it show on my gauge a change in vacuum and idle speed .
I'm running Hedman full length headers 13/4 primaries
I struggle with concept the carb is getting a bad signal, albeit possible. If you're dropping vac radically it's unlikely you'd make it to 3000 RPM without issue. Conversely, a spike in vac is unlikely with open throttle. But in both cases, you "could" have carb bypass issue (idle circuit, etc.), but that typically causes it to go rich, and not huge extremes. If the reading is going over 20 (22-24), the engine would be freaking out up to 3000 RPM. Back pressure is important for signal, but what you're experiencing is pretty radical.
Based on what you're sharing, I think you're dealing with a O2 placement issue. Also, those headers are plenty long enough, meaning the bung could be moved pretty far up. The "easy" fix is pop another bung in there and plug the current one. There's a min distance you want from the head exhaust port, but most people just say 6-8" past the collector. The really important part is the min amount of pipe you have after the sensor, which is 18-24". In a perfect world.. you'd have the sensor located 18-24" from the engine and 18-24" from the end of the pipe. Once you open the cutout, air is getting to the sensor from the long pipe, negating the length of the cutout extension.
I should have asked this first, but you're running a carb, not an EFI, right?
One last point, the more the sensor is exposed to the atmosphere when running (electrified) the more it shortens it's life.
OK.. one more comment. I would swap the sensor to the other side. It's an easy, quick check. it's unlikely that you'll see a difference, but at least that will exclude any valve issues in the right head. (it looks like there's a bung on the driver's side).