Easy peasy

[Chuck H]

Well all that said it sounds like this sensor just needs a steady 12V to operate. It's not sensing anything off the 12V line so I say go ahead and add protection for the sensor. A 12V/5W zener in parallel with something like a 100uf/400V capacitor across the sensors +/- terminals and maybe a 47ohm/10W resistor in series. Check voltage at the sensor after installation since I'm spitballing the added resistor. It may be a tad low of 12V but I'l bet it would work alright and help extend the life of the sensor. This circuit could be pakaged in a shrink tube and tucked away pretty easily I think. But it's your BBQ.

[g1]

You would have to decouple it from the ACC line, which should be coming straight off the rectifier. Although it is called 12V, it is fluctuating between 12 and 14V, so it should not be clamped directly. Unless in these modern bikes the ACC circuit is already decoupled from the charging circuit and regulated.
The battery acts as a big capacitor, so it should kill any big spikes.

[Pixel]

Before you could buy LED bulbs for cars I kept making myself some and they kept dying until I stuck a small film capacitor and 18-24 V zener across it. YMMV.

[Chuck H]

Before you could buy LED bulbs for cars I kept making myself some and they kept dying until I stuck a small film capacitor and 18-24 V zener across it. YMMV.

This is a good observation. g1 points out that regulating below the charging voltage may be problematic but the voltage spikes damaging the sensor may be somewhat higher. So spike control aimed at a higher voltage may be enough to save the sensor without interfering with the charging system

[Mick Bailey]

Bike alternators usually output quite a high voltage (I've measured 30v to 60v without a regulator or load) so are quite capable of generating damaging spikes if there's a problem somewhere. Spikes from the ignition system can also back-feed onto the low-voltage side. A bike of mine had a 'fix' under warranty from the importer where they spliced in a 1N4001 diode reverse-biased across the ignition module (on the input from the pickup on the crank). This was because the ignition was getting a sharp spike from the collapsing field from the ignition coil which was giving a false trigger and spurious mis-timed sparks. The fix was just the same as a diode across a relay.