It's likely more about the devil you know vs. the devil you don't. The backend software that banks, airlines, utilities (my industry) use have been vetted for decades (some for close to 50 yrs). The biggest issue with flipping to a new system is not the cost of the upgrade but the unknowns. It many cases it's lower risk to improve the existing system than start at v1.0.

Related to the incident. I agree with others that the true cause will take days. However, if GA Power's statement is accurate a switchgear failure does make sense and, in many environments, the switchgear is a hard to avoid single point of failure. Unless you have redundant wires going to the systems there has to be a point were you switch between primary and redundant systems. This is often done at the switchgear where you are switching the load between the grid and a back-up system(s). It's a fairly simple piece of equipment but the recovery plan is often having a spare to swap in. Based on the fact this was solved in hours and not weeks it could very well be that is exactly what they did. But, it could also be something entirely different.

Finally, having lived as a United flyer (GS) during the heat of their Intergration (that included several computer meltdowns) I am VERY impressed with how Delta has managed this. The only thing I got from UA during these sorts of incidents was a health serving of attitude (search UA's forum for "meltdown" for a sampling). They didn't give out food, didn't provide any real public statements, no CEO same-day video, no vouchers, nothing. My sense here is DL is really trying hard to get things back on track and keep people reasonably happy.