Well, first you would have to have a television screen that fits within the bounds of the Wiimote and the sensor bar area. So for instance, if you are pointing at the top of the television screen, the Wiimote has to have a good view of the sensor bar. This has to work for the top of the television screen, the bottom, the left side, the right side, and each of the four corners. While pointing at any of those places, the Wiimote has to be able to pick up the sensor bar to be able to triangulate its position. If any of those positions don't work, or if any positions within those bounds don't work then the Wii remote will lose sync with the sensor bar and you won't be able to aim at things in those areas.
If this is successful, you will have to map out the TV screens area digitally by recording each of the four corners of the TV screen. This would require you to get all four corners with respect to your position, the tilt of the TV screen left or right, the tilt of the TV screen up and down in relation to the Wiimote, and how far each corner is from the Wii remote on the "Z" axis, etc. Now the Wii would have to triangulate where the Wii remote is with respect to the sensor bar, where that position is with respect to the four corners of the TV screen, how much tilt there is on the TV screen, how far up or down the Wii remote is from the TV screen, the distance the Wii remote is from the TV screen, and the angle that the Wii remote is pointing. I'm pretty sure I left something out, but if that doesn't make the argument, then I don't know what will. I'm not going to even attempt the math.
edit: Just to make sure we're on the same page. I'm talking about lightgun accuracy. As in there won't be a pointer on the screen because you will point the Wii remote to your TV and it will work.