This is a very good question. The practical issues with antibiotic resistance may not be that obvious to a layperson because they tend to pop up in the hospital setting as opposed to the clinic setting that most people are familiar with.
The antibiotic bottleneck centers around patients who show up to the hospital with very vague signs of infection (fever, sweats, rigors, lethargy, etc.) and are sick. Like on the verge of going to the ICU sick. Or already in the ICU sick. These people need an effective antibiotic and they need one fast. You don't have the luxury of sending a culture to the lab and waiting for the antibiotic susceptibility profile to come back because that can take multiple days. So you need a toolbelt of "broad spectrum" antibiotics that are highly effective against the dozen or so bacteria that are the most common culprits of serious infection - something that you can use in the critical first 48 hours, while you're still waiting for culture results.
In practice (at least in the US), your selection of broad spectrum antibiotics is limited to vancomycin plus either cefepime or combination piperacillin/tazobactam. That's it. There really aren't very many other antibiotics that are effective against a sufficiently wide range of bacteria. We do have "back up" antibiotics to use if those fail, but these are powerful agents that come with some really nasty side effects, like seizures (for carbapenems) or widespread muscle breakdown (for daptomycin). And they are pretty limited in number too.
If vancomycin, cefepime, or piperacillin/tazobactam resistance starts to become a lot more common before the next generation of antibiotics comes in, we're screwed. And antibiotic development has been stuck in a rut for decades.