First of all I choose to believe that Talleyrand is not trolling at all because he hasn't said anything inciendiery and seems just to be speaking (typing?) his/her mind.
I think it's rather simple in the short run why they would keep prices high. Shrapnel knows its audience. The turn based strategy market is small, but I'm gonna take a leap of faith and say that it is older, which means it has more money. The kind of person who is going to buy this game is also the kind of person who is willing to pay a premium to fit their specific needs. In other words, lowering the price to go after a more casual gamer (in other words a gamer that is less inclined to buy this game, evidenced by them not buying at the higher price) is an illogical thing to do. Think about it: the more casual gamer that would be attracted by a lower price is exactly the kind of person to be put off by the graphics and complexity. In other words you're lowering your price to chase after a market that isn't there. In the meantime the folks that were going to buy it anyway at the higher price after being introduced to it are giving you less money.
There is no reason for them to simply follow suit with retailing practices for mass-market games because they are not selling mass-market games. They have no pressing need to clear inventory like a retailer (not to say they have no need at all to clear inventory but it isn't nearly the situation with a retailer where old product competes with new product).
There is also the following problem which has been alluded to but I guess I'll illustrate. Let's use some made up numbers and say that for the first 10000 games printed, the cost of the game to Shrapnel is 30 dollars including manufacturing. If they charge 40 dollars they make 100000 dollars in profit. If they charge 55 dollars they make 250000. That's hugely substantial. So that explains why they would start out at 55.
Now since this is software adding more copies sold decreases the cost per unit fairly substanially when compared to more "hard" goods like say a television. However Shrapnel, the folks with the data on sales and on their customers, have determined that the increase in sales for a lower price point doesn't make up for the loss in profitability even after the initial rush. If there is a market of 25000 copies at 40 dollars and a market of 15000 copies at 55 (assuming a cost of 30 per unit) it doesn't take a rocket scientist to figure out that you come out ahead charging 55. In fact the gain is 125000, a number which probably offsets the average cost decrease (the assumption of 30 per unit is not valid because the cost per unit is going to be less for the 25000 than for the 15000). Now these numbers are all made up but the point is that Shrapnel has numbers that *aren't* made up.
Of course the counter-point to all of this is, considering the steep marginal cost decrease once you recoup development costs, why not try and cash in? They lost money not charging less for Dominions 2 leading up to Dominions 3 if they had *any* inventory left, because the sales of Dom 2 after Dom 3 have got to be practically non-existent. They don't have any inventory left of course so that is a moot point V

V