I have a slightly different, though similar perspective on what would be the point of mirror go: Typically, you get a slight advantage from having the initiative in a local sequence, i.e. if both sides play the same number of moves alternately in the same region of the board, locally, the player who comes first should get the slightly bigger piece of the cake. Of course this is only small difference, but small differences can add up.
To Illustrate this, have a look at the following joseki:
- Click Here To Show Diagram Code
[go]$$B
$$ --------------------
$$ . . . . . . . . . . |
$$ . . . . . . 4 . . . |
$$ . 6 . . 2 . . 5 . . |
$$ , . . . . . 1 . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . 3 . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ , . . . . . , . . . |[/go]
Modern AI would play differently for white, but humans considered this even for a long time even though both players played the same number of stones(3) and black secured just a little bit more points than white with them. To me, this is explained by the fact that black deserves a little bit more just for being there first. It also goes in line with the common wisdom to first take empty corners, than approach others. If you play a game having black take the first, white approaching, then this joseki finishing in black sente, then black taking the next corner, white approaching and so on, unsurprisingly Elf gives 88% for black after the joseki is played in all for corners.
what does this have to do with mirror go? Now white gets komi, because black plays first. What does black get? He doesnt get more corners, in which he can have the first move, bacause 4 is even, which mans the 4 coners are miai, both get to start in 2 of them. The advantage of playing first comes now: black gets to approach/invade(in newfound AI 3-3 style), which is generally something considered to be "moderately sente", i.e. something you want to react to. Then a joseki of some kind will be played. You might say, what is the big deal, now white approaches one of blacks corners and they a similar joseki there with colors reversed.
This is true if white gets sente from the first joseki, but also black might get sente and use his next move maybe to protect his corner, so white doeosnt get to approach it in a state, where there is one black stone , but there already are two. And there it is the microscopic gain of playing first in a corner. In the same way, It is also a slight bit better for to have b-b-w-b-w-b-...-w ending in white (so locally one stone more for black) than b-w-b-w...-b ending in a black play (also locally one more move for black), i.e. being the "first" to play the second move in a local situation.
Which means white might give up just a little bit in order to take sente in the first corner black approached, so as to be able to approach one himself and so on..
Now another idea for white would be to say: I dont think your approach is sente. Yes, locally you threaten to do something, but I`ll just approach another corner and threaten something of similar size. That way, I preempted your attempt to be first in both places. But, if white does that he needs to make sure his threat really is as big as the other one, so that mutual damage would be okay for him.
Mirror go does just that: answering blacks initiative move with another initiative move of the same size: its mirror image. Of course this works only as long as the battles are local. Once they spread out to the center, white needs to be careful, because the "you do something here, I do the same thing to you there" concept doesnt work, once black gets to do things first, that help in BOTH fights.
I personally have problems against mirror Go with black in the 9x9 games I sometimes play on Go quest. I would be less scared on 19x19, cause I have much more time, to plan out, how I can drag the battle in the center, while on 9x9 it happened to me multiple times, that I took the left side and white got the right one, and the center is dame - I lose.