Just to elaborate a bit more, the problem is that even a colony which, for example, glows under a blue light, will have some members that didn't uptake the genes. So you need a more proactive filtering method.
There's also the fact that bacteria tend to lose non-adaptive traits pretty quickly once the selection pressure is gone, so having the antibiotic resistance attached to your gene of interest makes it much more likely to stick around for the duration.
Additionally, as others have said, this particular antibiotic (from the post) isn't even used for human treatment so it's not really particularly more "dangerous", we just tend to get nervous around anything "antibiotic resistant" because we jump to thinking about like, MRSA.
There's also the fact that bacteria tend to lose non-adaptive traits pretty quickly once the selection pressure is gone, so having the antibiotic resistance attached to your gene of interest makes it much more likely to stick around for the duration.
Additionally, as others have said, this particular antibiotic (from the post) isn't even used for human treatment so it's not really particularly more "dangerous", we just tend to get nervous around anything "antibiotic resistant" because we jump to thinking about like, MRSA.