Cleaning/replacing stuff?

[serialk11r]

Today I sat and idled for a minute or so (with warmed up engine) and looked at the data from Torque. It's not giving me commanded AFR or measured AFR, but I can read all the O2 sensor voltages. O2 sensors fluctuate between 0.1V and 0.8V readings.

Yesterday I discovered that my "load" (not absolute) was maxing out in the 70s (throttle pedal floored), which worried me a little. Car seems to stay in closed loop unless I lift and it goes into DFCO, even when I floor it.

As far as non-quantitative data goes, my car has a very strong unburnt gas smell when starting. After a bit of a drive, when I park and get out I feel like I can smell a bit of unburned gas too. When I had the car smogged, it passed with the maximum allowable amount of unburned HC (69ppm), and was emitting quite a bit of CO as well apparently. The only thing I noticed was that while the guy was running my car on the dyno the unburned gas smell was pretty strong, whereas the the previous guy's Camry smelled much cleaner.

Additionally, my idle is a bit rough and I notice it fluctuating sometimes.

I plan on cleaning the MAF next week, and checking the filter, but does it sound like an O2 sensor replacement is in order soon? The really strong unburned gas smell really concerns me. This tank my fuel mileage has been high 20s I'm estimating, and I did a lot of the miles on a cold engine so I don't think the fuel economy is being affected too much.

EDIT: Okay just went to log the car idling, and the results were disconcerting. Bank 1 sensor 2 is reading 0V most of the time, dead O2 sensor? D:
Bank 1 sensor 1 seems to be functioning, but the fluctuations are quite big. My scanner has 1Hz resolution (yea it was 11 bucks off Ebay whatever) but it shows oscillations between 0.85V ish and about 0.09V every few seconds. I read that with narrowband sensors the ECU is supposed to try to target 0.45 or something.
No idea what to make of Bank 2 sensor 1, it looks more normal than bank 1 sensor 2.

Okay nevermind, bank 1 sensor 2 is the post-cat sensor so it probably reads funny. Still, I feel like it's weird that the sensors are not locking onto a consistent value. Perhaps I should log it on a longer trip next time.