I think in theory that any process could make an adequate weld on any given application. Sure some processes are much better suited for certain applications, but they are not the only process that you could use to get the required weld. This just comes down to proving out how successful any given process is for the application. That's where procedures and procedure testing comes into play. You can do it however you want, but you better be able to produce records that show the success level is adequate per the job requirements.
I'm always open to looking at different ways to get the job done faster, but just because someone says they've done it that way before doesn't mean it will work for you.
BTW Lawrence, good analysis on the cost of over welding. That's a daily battle we fight here, and the welders just have a hard time to understand the overall cost impact.