I'll give my standard suggestion which is: hard problems, solvable using techniques your kids already know, are FAR FAR more valuable to mathematical development than learning new techniques (i.e., doing the next chunk of curriculum). That's what I'd be pushing for.
We did something like that in my second grade class, where every morning we had to copy addition/subtraction problems down from the chalkboard with increasing numbers of columns and rows, and it nearly killed my interest in math. Yes, they were more challenging, but it was still the same boring process we'd done a thousand times.
With gradually increasing frequency, the teacher would grade morning math papers and find mine missing. When she'd ask me for it, I'd defiantly tell her I didn't do it. She'd ask the girl that sits next to me to dig through the mess of my desk and find the paper I'd started all wadded up. Then she'd make me smooth it out and finish it.
I was a bit of an apple-polisher in school, so acting out in this way was a HUGE exception to my norm. And it's not like I had no interest or talent in math, because I participated in and won awards in district math competitions in junior high. It's just that doing the same thing every morning, even with increasing complexity, was more than I could bear.
My point is that enrichment and acceleration are actually two very different teaching strategies, and I'm not sure that, in the case of math, enrichment is a useful approach.
