Hmmmm trying out ideas doesn't sound very confident, because if you're just trying things out then you can wait a bit longer for compiling/rendering. XD
If something takes 10 hours, what's another couple of hours going to hurt? (Do things even take that long for most tasks in day to day use?, I don't have a clue about software development =P)
I'd also imagine 6 cores screaming along at 4.2ghz+ might be faster in most(all?) tasks than 10 cores at 2.3ghz, especially for things that aren't well optimised for multi threaded performance.
But even stock, I'm not seeing a big difference, have look at some benchmarks
here. Even if we cherry pick there are only a few things where the Xeon pulls ahead by 30%, I don't think that's worth an extra $1000 or more.
Taking all that into consideration, it just seems like pure
overkill waste and is a step backwards in some respects. I think it's better to go with the 5820k and divert the funds into more important things, and you're gonna have to sell me on getting workstation class graphics cards, what exactly are you using them for again? (next gen is around the corner, better wait for them because it's quite the process jump)
I've learnt the hard way that stockpiling unnecessary stuff is reaaaal bad. Cash flow is your highest priority and every single dollar that you spend must have a clear purpose into directly improving it. If you haven't made a cent using your PC/Laptop/whatever then it's just an expensive toy, and you should view it as such.
Like mechy said going for used Xeons is a very enticing option, because in my eyes it's only being used for hobbyist purposes so duck and weave all you can!