As far as I am aware, there are two ways to construct the five modes of the pentatonic scale.
The first is to go to the major diatonic, remove the fourth and seventh scale degrees (as these are a tritone apart) and call this the major pentatonic. Then just permute the formula. (Example: major pentatonic is {2,2,3,2,3}, next mode is {2,3,2,3,2}, next is {3,2,3,2,2}, etc.
But the second, and I believe the most fundamental, way to do it is to visit all 7 diatonic modes and remove whatever scale degrees create a tritone between them. For example, C Dorian contains [C D Eb F G A Bb], and the distance between Eb and A is a tritone. Thus the Dorian mode of the pentatonic scale (in key of C) is [C D F G Bb], which in terms of the semitone formulas is {2,3,2,3,2}, which is of course the first permutation of the major pentatonic formula.
When you repeat this analysis for Lydian and Locrian modes of the diatonic scale the pattern breaks.
My question: How do we interpret this?
For example consider C Locrian, which has [C Db Eb F Gb Ab Bb]. Remove the tritone between C and Gb, so we're left with Locrian pentatonic of [Db Eb F Ab Bb]. This obviously doesn't contain the tonic (which makes it unwieldy to use). But in terms of the interval formula this is {2,2,3,2,3}, which is the major pentatonic!
(Similarly for Lydian, you end up with the same formula as Mixolydian pentatonic, but with no tonic).
As there is no intervalic difference between Locrian pentatonic and major pentatonic, does that mean Locrian pentatonic doesn't exist?