I might not be as active for a while, but here's something I've been pondering about...
Do you prefer slow or fast regression? And by that, let's assume a regression that doesn't alter the clothing (so it gets loose), from fully-grown adult to about toddlerhood (3-5 years old). So a slow regression would illustrate before, during (e.g. teenager, preteen, child), and after, whereas a fast regression would just have before and after.
Here are a couple of pros and cons I can think of.
+ You get to see the person slowly realizing what's happening to her.
+ Gradual loss of clothing, e.g. from well-fitting bra to slightly loose bra to a bra that's way too large and needs to be held up.
- May lose 'shock' impact, i.e. we just want to see the end result anyway.
- More work, and takes more time to complete.
+ Maximum shock impact, the person would barely have time to register what just happened - reaction could be priceless.
+ Less work, no need to keep redrawing the same clothes on a slightly smaller body.
- May be too fast, everything would come loose all at once, no time to appreciate each individual change.
It's really an argument between having less or more images I suppose. In my opinion, absolute minimum is one, which is the end result, and easiest to do. However (unless it's an established character that most people are familiar with), it might be more impactful to show the before (fully grown adult) version as well for contrast.
So the question is whether the in-between stages add to the piece, or is just a waste of time since in the end all we want to see is the end result. I think if it highlights a particular change that isn't present in the other images, then maybe it's worth it? What do you think?