My Volvo is a bifuel variant so the ECU can switch between regular gas and compressed natural gas, after reading a bit on the subject it appears that CNG needs a higher Lambda to the tune of 1.15 or ~17 AFR and that since these fuels are already evaporated, rich running causes higher EGT.

In my Volvo I noticed that the oxygen sensor AFR was 14.7 for CNG, doesn't this mean it constantly runs rich instead of stoich? Even during idle.

Is there a particular reason why I shouldn't aim for stoich(~17) for fuel economy and less EGT? Was this done for the purpose of getting the power target? But CNG also has a higher resistance to knock so it also needs more timing compared to gas, right? But yeah for now I think I want to focus on whether this is too rich out of the box.

Please ignore what I said, I should've looked at it from the perspective of Lambda, and lambda was 1.0 in CNG mode. It was my erroneous formula that led me to view the AFR in a different way.

Perhaps I can rephrase the question to "Does a wideband sensor detect correct stoich for different fuels?"

The wideband sensor reports lambda. That's why it's called a lambda sensor.
To get the AFR, multiply the lambda by the stoich ratio.

If you have an aftermarket wideband then that will be usually always set for gasoline and show always 14.7. Because it just takes the measured lambda and braindead multiplies it by 14.7.
If you can, just set it to lambda mode. AFR is an absolutely useless metric.

The difference in leaning out the mixture is minimal, and of course there are issues with emissions.
Mostly what you are reducing past stoich are the pumping losses of the engine. Because going leaner than stoich in total you need to inject roughly the same amount of fuel anyway to get the same power.

If you can run lambda 1.05 from 1.00 it does not mean directly that you now use 5% less fuel.
Because to get the same power you have to get 5% more airflow to still have the same amount of fuel injected. So in effect the fuel usage stays the same, but the amount of air going through the engine increases and the amount of vacuum in the manifold decreases.
What you are gaining is the difference in pumping loss between the two situations.

This is what most people don't understand.

When running rich because there is not enough oxygen for a complete burn you are just dumping a bunch of the unburned fuel out of the exhaust.
But at stoich you have more or less enough air to oxidize all the fuel. So all your further gains are from reducing pumping losses.
Of course this all depends on the atomization and the combustion chamber design, but in case of CNG we can assume that the atomization is nearly perfect.

Will running e.g. 1.05 be more fuel efficient? Sure! Will the difference be more than perhaps 1%? Highly unlikely.