When I wanted strings in my app to be translated, I included screenshots so that the translator would understand the context. Otherwise it might be impossible to guess what a very small string is supposed to represent.
For a while I was working as a manga translator, and in that case it would have been doubly important to see more context. Actually it probably wouldn't even work to split a translation task for multiple translators, since each might decide to translate something inconsistently. For example speaking style, naming of things etc.
This will probably work best for longer pieces of text like business text where you can easily guess the meaning without seeing any context and where you don't need to make decisions that might affect text that comes a hundred pages later.
This is a very interesting startup, if you want more gyoza distractions I'll be travelling in Tokyo until 3rd of May.