Nah, you prolly can make it work with vectors in ecoli but this is too flimsy and hald-assed... maybe worth trying as a proof of principle but I wouldn't put my money and effort on this route. Depends on what you want finally achieve though, that is do you want to do MO for your joy (where you can grown ecolis with your lab antibiotics and lab media) or you look for something for the bigger audience (i.e. useful and easily usable for the community). I'm for the latter, hence I vouch for the yeast and genomic integrations.

BTW, you're the first to claim ecoli smells nice, to everyone i've met it really does not compare with the bready sweet smell of yeast.