Usually the capacitance values are chosen high to avoid ringing on supply lines and/or set the resonance frequency as low as possible. It may also be the lazyness of engineers: "0.1 uF should be enough ...". Ihave seen 47n capacitors used for decoupling but lower values .... never .... Snubbers may also be used to eat up the ringing.
Since when was it laziness to concentrate on the important stuff?
I always use a "grab bag" value of 100nF -20/+100% for decoupling caps, no snubbers or similar, and have NEVER had a problem with "ringing" or "resonance frequency". That the idiots that do not use ultrashort traces for the decoupling caps may have problems has nothing to do with this. I do, however, give the uC (just the uc) an additional 10uF tantalum.

For that same reason one should not make the decoupled loops identical but slightly different. One way is to use two values for the decoupling capacitors.
Beware of snow in July. It is just as likely as it is that the above makse sense.

Ihave seen 47n capacitors used for decoupling but lower values .... never.
My experience is that 10nF is quite popular.

Erik