RadioBanter - FM Channel Spacing vs. Carrier Bandwidth vs. Audio Response

I'm not an amateur radio operator, but as an SWL, I figured that this
would be the perfect newsgroup for posing my question.

Every slice of the radio spectrum has a designated frequency spacing.
200 kHz on the broadcast FM band (88-108 MHz), 12.5 kHz on the UHF-T
land-mobile band, 25 kHz on 2 meters, and so forth.

Is there any concrete relationship between said frequency spacing and
FM carrier bandwidths? In other words, would I be correct to assume
that a commercial FM radio designed for 506.5125 MHz will always send
out 12.5 kHz-wide carriers, while a 2 meter amateur rig's FM carriers
would always be 25 kHz wide?

The reason I'm curious is because of what I experience when scanning
all these bands with my Rat Shack PRO-2006 In NFM mode, this scanner
copes with ALL narrow-band FM carriers regardless of whether they're
12.5 kHz wide or 25 kHz wide. Assuming that its NFM tuning circuitry
were adjusted for a median carrier bandwidth of, say, 18.75 kHz, I'd
almost expect to hear some amount of overdeviation on 25 kHz carriers,
while hearing some amount of underdeviation (high SNR, low audio) on
12.5 kHz carriers. But that doesn't happen.

What's the explanation, then? Are receivers like these simply capable
of "coping" with carriers slightly narrower or wider than their NFM
circuitry's "median" bandwidth? Or do all narrow-band FM carriers
simply exist at one common bandwidth regardless of channel spacing?

My inclination is toward the former: the carriers' bandwidths vary by
available channel spacing, and receivers like the PRO-2006 just "cope"
given the very slight difference between 12.5 kHz and 25 kHz bandwidth.

Which is it? :-)

Finally, and assuming that my inclination above is in deed correct,
is there any hard rule for figuring out the potential maximum audio
frequency response of an FM carrier assuming that you know said
carrier's bandwidth? While mulling over the specifications for some
commercial Motorola radios, I noticed an interesting coincidence.
All radios designed for operation in the VHF band (25 kHz spacing)
were quoted as having 5 kHz deviation. OTOH, those that were meant
for frequencies spaced 20 kHz apart had 4 kHz deviation. And those
designed for operation on the UHF-T band (12.5 kHz separation) had
2.5 kHz deviation figures quoted.

There's a pattern there. Deviation = 1/5th channel spacing. But is
this really how one determines the maximum deviation for FM carriers?
I can't help but noticing, for example, that FM broadcast radio has a
200 kHz spacing, while the maximum deviation permitted is 75 kHz (the
portion between 15 kHz and 75 kHz being used by pilot tones and stereo
difference signals, etc.) Yet 1/5th of 200 kHz = 40 kHz.

The usual apologies for a long post, and thanks in advance.