What I am interested in, however, is ideas on how to disrupt/fix enterprise software via startups.
The previous developers who put together the code aren't totally clueless people. Unless we wish for our own work to be brushed off in the same way we did to our predecessors. Future developers are going to laugh off at the cheesy way we develop our mobile and tablet apps.
Some of the original code had to work under onerous memory constraints, and have all manners of workarounds for various platform incompatibilities. Think IE6 hacks, pre-AJAX code when a sizeable portion of browsers could not script.
Do I think enterprise software as it is today can be disrupted? Yes, I do. But do I think the enterprise software vendors and developers are clueless? No, absolutely not. Many of them have adopted agile processes and are as fast as any other developer. If you want to disrupt enterprise software, you need to disrupt the sales channel, like what Salesforce and SuccessFactors did.
Go on, I am very interested :)
Furthermore, a lot of these large companies start off with a MVP hastily knocked up from 4GL tools and spend a lot of time doing customer development. (remember lean?)
Eventually they get big enough to be able to afford proper software engineering processes.
Gianforte of RightNow, a big enterprise player, started off in his bedroom with a list of possible features and lots of phone calls.