It's optional but we do it and believe in it too. So does Collings and since the last time that this came up on the OLF David Farmer educated me that Taylor does it too.
It's important to understand what the benefits are before deciding if it's for you or not. Here's what I like about the idea of the rabbet on the bridge bottom:
1) If you have ever removed a f*ctory guitar bridge you may have noticed that f*ctories can leave a LOT of finish under the bridge eliminating wood-to-wood contact. It's often one contributor to why some bridges lift especially when the wood-to-wood contact is not good on the back edge of the bridge where the heavy lifting is done in keeping it in place.
Using a rabbet can and does in our repair world greatly expand wood-to-wood contact and I've even measured one example where the maker, a f*ctory, left 40% of the available gluing surface under thick finish.
With this said the major functional benefit of using this technique is more gluing surface on a critical, under constant tension component while preserving perfect cosmetics around the bridge.
2) No one likes the look of a bridge where the finish has been cleared to the perimeter of the bridge where with time, RH swings and if the wrong glue was used and there is creep the cleared finish area becomes visible for all to see.
Additionally repair folks may hate a guitar where the finish was cleared to the perimeter because it will often require some level of finish touch up if the bridge has to be removed, or lifts..., and reglued. We like our repairs to be as good or better than new.... and this means invisible.
Although clearing finish to the perimeter is as much wood-to-wood area as possible a .050 rabbet is really mouse nuts in terms of losing gluing area.
With this said we prefer the rabbet over clearing to the perimeter.
3) When you rabbet the bridge you also clear the finish to match and create a "pocket" of sorts where the bridge will "snap" into the well that we created. What's great about this is when using HHG, you all use HHG don't ya.....
..., and time is limited being able to slide the bridge into a resounding detent and then sung up the clamps just makes one's day! It's also beneficial in not gluing the bridge on in the wrong location and also stops the bridge from sliding around during the application of clamping pressure. The bridge rabbet is part of a system and in my view needs to be thought of this way. Although you can do it with Titebond with no issues it really shines with short open times such as HHG.
For you guys your finish thickness should be predictable. You know how much finish you shoot, presumably or often have experience with prior instruments and you should be able to closely predict finish thickness. When you can't, such as in my repair world, measuring finish thickness is easy and that's what the depth is set to, what we measure.
One last thing if I may please.... When I was building my goal was never to approach the quality of the best f*ctory instruments. That bar is pretty low in my experience.... Instead as a custom builder I always believed that our value add was to take it to new levels, do all the tricks, hot rod the thing so that the value proposition is "superior" to that of a f*ctory instrument. Being sure to build in "serviceability" for extended life AND doing all that we can to add value to our instruments was in my view what this is all about.
Rabbeting bridges is one small and easy thing that we all can do to help ensure that folks like me won't be billing our mutual clients when the guitar that they bought from you has a bridge failure.... It still can happen mind you but if you take advantage of all but say .050" of bridge perimeter as available gluing surface your bridges will likely require more trauma to make them fail.
Can a great guitar be built without rabbeting the bridge - sure and they have been for over 100 years now.
I believe the rabbeted bridge and the recessed pocket to bare wood is the way to go.