It may depend whether you’re trying to maximize awesomeness and achievement or to minimize financial penalties. Let’s aim for some of both. There’s two legit ways to stay on a goal’s road: do the thing and make the thing easier to do.
I definitely noticed that my high pledge goals gravitated toward the latter, by having more conservative slopes and less ambitious fine print. For me, that’s a net badness.
Part of the reason that we introduced pledge caps, iirc, was an attempt to stop the side-effects of monotonically-increasing exponential pledge amounts.
A good example is a mustdo goal, where each day you specify the task to be done. Eventually, your task becomes ‘get out of bed’ and the goal isn’t worth very much any more. It’s a less visible slippery slope than a rising pledge amount.
An automated down-pledge might counteract the tendency toward awesomeness-reducing conservativeness. Yes, it might have a financial cost, but on balance I think we will create higher quality goals as a result.