Let's see if I can break this down a little more...

I believe in going the extra mile. However, in going the extra mile, would'nt you consider the most efficent way of getting there? i.e., if the problem should only cost $100 to address, why should I exert $500 of effort when I can exert $100 of effort and achieve the EXACT SAME RESULTS.

The argument here is about efficiency in implementation, not in robbing the client of features. From the client's POV, he sees no difference.

I think you have mis-interpreted my position. I believe in juicing up the grid as much as one can. The only rule I impose is that the underlying column order has to match grid column order. And by the way, this is ONLY an issue if the grid column order is subject to change. If the order is static, the issue does not exist.

Clear??