You can take your pick from the big 3 wideband controller makers- PLX, AEM, Innovate. They all work. The latest models use Bosch 4.9 sensors. We find AEM and Innovate are somewhat better than most other brands. No experience with the NTK based ones.
Bosch publishes guidance for sensor orientation and preferred EGT range here:
https://imfsoft.com/files/Lambda_Sensor_LSU_49_Datasheet_51_en_2779147659pdf.pdf
Just to be clear, all WBs have a heater element to hold the array within a specific temperature range to be accurate. If memory serves me correctly, this is not below 1450F. The sensor can only heat itself, cannot cool. The sensors can withstand up to 1800F for very short periods but life will be shortened. Thermal shock is a leading cause of premature failure due to condensation forming in the exhaust on warmup. Follow Bosch's orientation advice. It's important to avoid this issue. OEM controllers ramp up heater current slowly to avoid thermal shock. I'm not aware of any of the low cost WB controllers doing this. One reason few aftermarket O2 sensors last anywhere close to OEM application ones.
The sensors do degrade slowly with age. Bosch also mentions that TEL (lead) affects sensor life substantially depending on concentration and is likely to be 1/10th as long as operating under the same conditions with unleaded fuel.
We've seen them last (with drift) as long as 350 hours and as little as 2 hours on 100LL. Drift on may reach .6 AFR at 2000 hours on unleaded fuel according to Bosch.
Running AFRs richer than about 11 to 1 for long periods will shorten life.
I've seen some independent tests on a bunch of aftermarket WBs and many were not all that accurate, despite manufacturer claims, when compared against lab quality equipment. Some were downright inaccurate.
We've had 15+ years of experience with these things on thousands of EFI systems. 6 different controller brands. When the sensors are new, they can offer a good guideline for AFRs and tuning. When using leaded fuel, degradation may be pretty quick (in some cases). If you have EGTs, you can check WB accuracy by leaning to peak. On clean (no ethanol) mogas, the AFR should be around 14.7, with avgas, around 15.2.
On aircraft, there is the question of accuracy with altitude, you can check this aspect as well by leaning to peak at a few different altitudes. Indicated AFR should be close in all cases. There are some technical reasons with O2 pressures as to why they may not track with good repetition if calibrated at a different altitude. This may not be so important in automotive use but certainly could be in aircraft.
In short, don't believe these are the gospel without question. The only thing worse than not knowing what the true AFR is is to have a gauge you believe to be correct and isn't, while you chase tuning. Good to question these things armed with some understanding and logic. Too many people believe digital instrumentation must be 100% accurate. There are lots of reasons why this may not be so.