Generally boundary mounting means mounting the mics as flush as possible to a large surface. The key words there are flush and large.
Here's the two primary variables (described in terms of geometry without any math or specific figures involved, just the basic relationships)-
1) The boundary effect correlates the distance from the surface with the upper bound of the frequency range effected.
That means you get boundary layer effect up to a frequency which is directly related with how flush the microphone diaphragm is to the surface. Get close the mic close enough and that's effectively pushed up to around the high frequency range of human hearing or above. If the mic is not close enough you'll start to get some high-frequency comb filtering effects above the upper bound from reflections off the surface interacting with the sound arriving directly at the microphone diaphragm. In real-life terms that means the highest stuff may sound either a bit bright or a bit dull. It's sort of like introducing a wavy EQ curve at high frequencies. This one is easier to deal with in some ways- just mount the mics with their diaphragms as flush as possible to the surface. Laying them directly against a surface is usually close enough.
2) The effect directly correlates the area of surface to which the mic is mounted with the lower bound of the frequency range effected.
That means the boundary effect only works down to a certain frequency determined by the size of the boundary. If you want boundary effect at very low frequencies, you'll need to mount to a very large (and acoustically rigid) boundary. The shape of the sensitivity boost is like a shelf filter and the corner frequency of the filter shifts up or down depending on the size of the boundary. For that reason, boundary effect mounting is usually on large surfaces like the floor or a wall, or at least something "near human size" like a big desk or a 4' square piece of plexiglass or something. The sensitivity increase effect still occurs on small boundaries, but only within a smaller frequency range down to a higher frequency - a much smaller range down to a much higher frequency if the surface is much smaller.
In terms of the effect of the spacing between omnis and "stereoness", there is no difference if the microphones are in a boundary layer zone of a surface or out in free space. There will be a difference if instead of having both mics mounted on the same surface, they are mounted to separate surfaces with an angle between them. At that point they begin to act like directional mics within in the frequency range effected, and introducing more angle between the two surfaces can make up for using less spacing between the mics. Using two pieces of plexiglass (perspex for any Brits reading this) hinged in the center to form a wedge shape is a relatively well known stereo boundary mounted technique. But each sheet of plastic is usually something like a square meter or so to get a reasonably low frequency bound on the effect.
Likewise to angle, a barrier introduced between two boundary mounted microphones works the same as using a Jecklin disk between two microphones in free space. If you have a barrier between the mics, you'll probably want to use less spacing. The bigger the barrier, the lower the frequency range in which the barrier is effective. A small barrier only does it's thing above the frequency which relates to it's size and is "sonically invisible" to sound at lower frequencies.
For a rather small boundary like a plastic CD case, you're talking around 2kHz or so (without doing the math). Mounted up on a pole, you'd get basically an response bump from 2kHz up. More like an EQ adjustment for sounds on-axis for the most part. That might be useful if you want that presence range sort of EQ boost, or if you want to angle the planes apart to try and get more stereo effect in the effective frequency range with an overly narrow spacing or something.
Oh, and like a Jecklin disk, the shape of the boundary very much secondary to the area of the boundary. It's mostly size that counts.
The real-world answer is to try it and see if it sounds like what you are looking for. Play around with it some more, and listen critically. Use the same mic spacing with and without your makeshift boundary mounts in the same location and make up your mind based on what you hear.