Even this strikes me as a bad idea. The extra time investment aside, what happens when the translator doesn't understand some of the core concepts going into a script translation? Stuff like line length limitations and section breaks. It's easy enough to explain, but even when these things are understood errors tend to crop up in the scripts. So the hacker sends the script back, the translator fixes a few things. They play ping pong with the script a dozen or more times before things really start to clear up, the hacker's completely fed up by the end of it, and when end-users complain of badly-formatted text the hacker's the one that gets it in the gut even though he probably doesn't even understand the language that his stuff got translated into.
Don't get me wrong, it's a nice idea in principle, but we're not living in an ideal world and I just don't think it's practical.
This was actually the reason we released the Mother 3 translation tools a while back (the Mother 1+2 tools too, forgot about those) and then said, "You can use these but don't expect support from us."
Originally we gave our tools to a bunch of skilled translators, limiting them in number so we wouldn't be swamped trying to give support to so many people at once - but then many of the people would disappear for various reasons (life, underestimating the amount of work needed, getting bored of the project), outsiders would show up asking for tools even though their language's translation was well underway, people would ask for significant hacking work that we didn't handle in our original translation, we'd have to do lots of back-and-forth whenever something broke (and things often broke)...
Having experienced it first-hand for years, I would NOT recommend the "let multiple translators translate the stuff then send it back to the hacker" method. It sounded like a good idea at first, but I've been much happier with the "throw stuff out there and let people fend for themselves" method. It's shown greater success, too.