Re: I hope you can fix stupid!
Tom
I think what might have worked would have been to join all the corners, then run a scallop over the dovetails so that just the dovetails would be reduced in height, all around, by the degree of the bevels. This if done elegantly would add an interesting detail. If I haven't got the right end of the problem, there are always ways around it. You can be sure that most decorative elements probably started up for some similar reason.
Another option would be to cut open and redo two of the corner's assuming Barbie would still fit inside.
The bigger lesson is to suck it up when mistakes happen and if necessary start again. All the work that goes into covering up an error, all the work that still remains to be done. And for what? To get something you didn't want in the first place. Something people are all going to have to pretend is "great" for the rest of time. It is part of maturing in a craft. It may even cause a scar deep enough that one doesn't make similar mistakes.
It has always been accepted that fixing mistakes is part of being a great craftsman. Even if one never makes mistakes they crop up in stock, or the work of people around one, or in repairs. But the objective of invisibly fixing stuff without breaking step should not affect results. If something goes wrong with my boat project, I can fix almost anything, but it has to be in a way where it isn't going to break way offshore.
People sometimes wonder why they can't shoot the arrow in a 10 ring consistently, or bag the big buck. Well there are only hits or misses, if you have a huge degree of satisfaction in almosts, that is what you get. If you don't pass on the little bucks, it will take an act of god to get you a trophy.
I am a frequent advocate for the idea of learning through practice not projects. This is really excellent advice, that I don't expect anyone to actually follow (at least not in woodworking). But when something goes wrong (and it isn't an actual eureka discovery of an even better outcome) you can write off the losses as just being practice. If a person won't do the practice then maybe the next best thing is to convert errors to practice. So ends the sermonizing.