Will Mcacc and/or Audyssey make baffle step compensation unnecssary?

Baffle step should be measured and not estimated. The theories used to calculate baffle step don't accurately reflect what occurs in a real placement.

Baffle step is function of boundaries with the baffle only one of them. If the woofer is near the floor or ceiling or if the speaker is near a wall you will likely need far less. The only way to get it right is to measure. Where will the speakers be placed relative to boundaries?

As far as whether the room correction in an AVR will adequately compensate for this, I can't say for sure. In theory, it should compensate but my experience with Audyssey says the results can vary greatly. Are you designing a passive crossover or active DSP?

If you want to push the baffle step to a lower frequency make the baffle larger. If you want to eliminate it flush mount the cab into the wall. Otherwise you must use BSC in the crossover to reduce the sensitivity of the speaker above the baffle step frequency, or EQ it to arrive at flat response. But BSC has an inherent flaw, as it doesn't consider Allison Effect losses when the baffle to rear wall distance is 1/4 wavelength, and rear wall reflection augmentation where it's closer than 1/4 wavelength, so I don't use it at all. So long as you have an EQ you might as well use it, but be aware that no EQ can compensate for Allison Effect, so you need to consider that from the beginning of the design process.

You didn't mention the baffle size or proximity to the surrounding boundaries.
Before it became common place to make such compensation a permanent function of passive crossovers - it was addressed ( when needed ) with an adjustment of bass.
My mains are in-wall and I wouldn't switch - but it's not practical for everyone.

I would definitely not use an undersized box...bad band-aid. As said before, design the crossover for the environment and/or use EQ.

In my experience, it doesn't do a good job compensating for a lack of BSC. It may just regard the speakers as small or LF limited and recommend a crossover frequency higher than desired.

What exactly is baffle-step compensation and how is it implemented in a crossover, both passive and/or active?

Marty - Here is a rather simplistic explanation of BSC:

http://www.salksound.com/wp/?p=42

Above the baffle step radiation is half-space; that frequency is where the baffle is one wavelength wide. Below that frequency there's a transition to full space radiation, which is 6dB lower on-axis than half-space. The crossover may be configured to attenuate axial response above the baffle step to realize flat axial response throughout the speaker bandwidth. But there's a major fly in that ointment, that being room modes, especially the reflection off the wall in back of the speaker. If the baffle is less than 1/4 wavelength from the wall that reflection in and of itself compensates for the baffle step. To get perfectly flat response in-room requires that baffle step compensation be performed with the speaker in place. We used to do that by externally mounting the crossover, allowing relatively easy component swapping. Inveterate tinkerers were known to do so for weeks, if not months or years. With DSP it's a lot easier to ignore BSC entirely and use the DSP to get flat in-room response.

I too used to diddle with crossovers incessantly before DSP. That was then, this is now. What could take months of fiddling to do with a crossover I now do in thirty seconds with an RTA and DSP.