(The Darkei Moshe in Hilchos Tzitzis discusses why the bracha we make on a talis kattan is "al mitzvas," which is a rare nusach for a bracha. He says that many people wear a tallis kattan that is smaller than the requisite measurement, and so the bracha might be a bracha le'vatala. To avoid this problem, he says, we use the nusach "al mitzvas," because then it can function as a general bracha of praise, a birkas shevach on the fact that Hashem gave us this mitzvah. If we would say La'asos tzitzis, or lilvosh beged metzuyetzet, it would be an unequivocal birkas hamitzvos, and absent kiyum mitzva, it would be le'vatalah.)
great unknown, in the comments, brings the Gemara in Brachos 12a,
In OC 209, MB SK 6, he says that although by brachos that are miderabanan we rely on the kula, and you are yotzei, by brachos that are de'oraysa we are machmir and require that he repeat the bracha, because, apparently, we lean toward the opinion that the kavana while he says the Sheim Umalchus is the main part of the bracha, and kavana for one purpose cannot be recharacterized.
So, the bottom line is, that the reason the Mechaber's idea and the Tzlach's idea work, is because the switched ending is consistent with the idea of bracha in general- lamdeini chukecha, or Dovid Hamelech's bracha of Hashem. But here, forget about the minor change from Object to Subject. We're trying to change it from Bracha to recitation of history. That kind of change doesn't work even toch kdei dibbur.
Yasher koach, great unknown. However, as great unknown points out in his comment, although the tzushtell to the Gemara is excellent, once you bring in the Mishna Berura who says you can't even repurpose a bracha de'oraysa once you've said the sheim umalchus, then we have a problem with the advice about Lamdeini Chukecha! Because if you can't change the kavana of the Sheim from Hagafen to Shehakol, you certainly can't change it from Shehakol to Lamdeini Chukecha..