Since Poker always starts as a struggle for the antes I think that implies that as the blinds go up, semi-bluffing should be more profitable. Combine this with the fact that you should play looser for larger antes, and tighter for smaller ones. I suspect that stone-cold bluffing would be more profitable at higher stakes too? (Often, semi-bluffing is a good way to start a successful stone cold bluff.)

Of course, the degree of profitablility is a function of the probability that your opponent(s) will fold. That is, the above conclusions rest on the assumption that you have to be able to reasonably suspect there's a possibility your opponent(s) will fold. So, I think if you are playing with a bunch of calling stations, maybe implicit collision would dilute the profitability gained by increasing your bluffing frequency, and maybe even ruin the edge you'd gain from bluffing more for larger antes/pots.