Everything was perfect before except for L and R being backwards.
Did you mean just L, or R being backward? If you had them both backward then the L and R and the center will likely cancel each other's low frequencies to certain degree while the cancelling effect with the surround speakers will be hard to tell due to the distance between them. If you only have either the L or R backward (that is, reversed), then there will be some cancellation effect between the L and R (and with the C too) in the low frequencies.
In any case, the "phase matching" feature probably acts like some sort of "loudness control", similar to D&M's dynamic EQ. In addition, as implied by the name phase matching, it probably has the ability to detect and correct issues due to one or more speakers being wired in reverse polarity.
Would that have caused the "phase matching bass" mode to behave in a different way??? This makes no sense.
If my assumptions above are correct, or partially correct, then yes, the "phase matching bass" feature should behave differently if you wired one or more speakers incorrectly. Once the wiring are corrected, the "correction" effect should be limited to that of something like D&M's DEQ, that is without having to correct the phase issue, hence less pronounced effects would be observed because there would no bass cancellation due to speakers being wired in opposing polarity to begin with.