The post culminates in a comment by BenJ:
Budgeting doesn't need to be so complicated anyway.
Amen! I think that's why we're all here.
Both Microsoft and Quicken employ some of the most intelligent software people on the planet, but getting a simple, easy-to-use software package out the door in such large companies will continue to be a problem until two issues are solved.
Issue 1 - Put together a small team and leave them alone.
Updating people and justifying decisions to a vast array of senior managers can suck up a huge, and I mean HUGE, part of a team's time. Once the project is determined to be worthy of funding and you have your crack team in place, put a trusted senior manager/executive in charge of monitoring them and leave them alone.
Encourage the team to be in contact with their colleagues on other teams so that they have insight into what other projects are doing, but it's hard enough to do this stuff right. Let them focus.
Issue 2 - Let users buy/build extensions to add functionality they need.
First, develop a simple, easy-to-use basic framework for the software and then create extentions to let users integrate functionality specific to their desired uses. This works well in basically any software (personal, business, or enterprise). For you business strategizers out there, this can be turned into an income stream quite easily.
In fact, go one better and let other people develop extensions. The oft-mentioned APIs (Google Maps and craigslist) that created housingmaps.com are very powerful in generating ideas for future development. If you find one you really like, license it. In this case, your customer works with you and gets something back. Heaven forbid.
As I work in a large organization, I feel the pain of the MS Money and Quicken teams. These tactics do not need to be unique to 37Signals and Adaptive Path. It seems like people are now frustrated enough, the tools are available, and the audience is starting to look elsewhere.
Life is complicated. Software shouldn't make it worse.