What would go wrong (almost forced to) is inevitably the creation of mosquito populations resistant to this bacterium or bacteria that infect other hosts than those mosquitos.
Any perceived correlation between Zika outbreaks (like 2013 French Polynesia, 2015-2016 Brazil) and those modified mosquito releases (like 2010 French Polynesia, 2011-2014 Brazil) is probably just pure coincidence. Even more coincidental is that ~15 years ago the islands 300 km from Gabon were specifically identified as the best isolated places to test the modified mosquitoes and the 2007 Zika outbreak in Gabon, one of the first major outbreaks (another one was in the same 2007 half a world from Gabon - on remote island Yap :).
Could it not be that the correlation is that geographical regions with higher incidence of Aedes Aegypti would have higher likelihood of having Zika outbreaks and would also have local governments more inclined to pursue mosquito reduction efforts?
