I just wrote and deleted a very different answer. Instead let me tell you the story of Lubos Motl's desk.

When I started graduate school at Harvard, I needed a desk for my new apartment. I found one on Craigslist and wrote the anonymized email address asking whether I could come pick it up. Who should write back, but Lubos Motl, who delightedly assured me that the desk was available, that he remembered who I was from my grad school application, and that he lived down the street. I headed over to his place.

Lubos was leaving his Harvard professorship to go back to the Czech Republic. After suggesting that I also buy his electric razor and his boom box (I declined), he asked what I was going to study in grad school.

For new high energy theory grad students (G-zeros), the big question is always "string theory or phenomenology". This is kind of an unfortunate question because new students usually don't know what either of those things actually is (I cringe to think about all the uninformed nonsense we G-zeros flung around). My knowledge at the time mostly came from the same place as OP's: I had read many debates online about whether String Theory was real science, or whether String Theory was dead, and I wasn't sure. Mind you, I had actually taken Harvard's graduate String Theory course as an undergrad, but the course was so detail-focused and I was so confused that I really didn't understand the big picture.

So Lubos was asking me this question, while helping me carry his old desk down the street, and I reluctantly admitted that I probably wouldn't study String Theory because I didn't understand the point and it seemed weird. He was... disappointed. He left the desk on my doorstep, and that's the last I saw of him.

So I started grad school a skeptic, swayed in large part by the Woits and Smolins of the world. I chose an advisor that didn't work on String Theory and convinced myself that phenomenology was better for me because it was closer to real science.

Here's what changed my mind about whether String theory is worthwhile. When you do math and physics, you develop a strong appreciation for those 'wow' moments when a calculation works so spectacularly beautifully that it MUST be hinting at a deep and magical structure that MUST be true (think e.g. Langlands program, not collider predictions). As I learned more about String Theory, I got better at understanding its 'wow' moments (of which there are many).

Here's my personal favorite at the moment (Warning: technical jargon -- feel free to skip this paragraph). Physicists have recently been able to compute operator dimensions and scattering amplitudes in planar N=4 Super Yang Mills theory nonperturbatively at all values of the 't Hooft coupling. These quantities interpolate between 5-loop Feynman diagram calculations at weak coupling (note, these are calculations in QFT, a theory no physicist -- even Woit or Smolin -- disputes the correctness of) and 10-dimensional supergravity calculations at strong coupling, with full String Theory calculations in between. This is spectacular evidence that N=4 SYM (a close cousin of the theory of quarks) is both a Quantum Field Theory and a String Theory (as conjectured by Maldacena), in a way that can be made quantitatively precise.

I don't know what high energy theorist isn't impressed with 5-loop Feynman diagrams. And if you're impressed with that, you'd better be impressed with all-loop results that agree with 5-loops when Taylor expanded to 5th order.

That kind of result is so impressive that I am convinced without a shadow of a doubt about the merit of studying String Theory. It also shows why String Theory is an inevitable and inextricable part of physics. Gauge theory is real and correct: it describes electrons and photons and W-bosons and quarks and has been verified innumerable times by a variety of experiments. The above evidence shows that some gauge theories literally are String Theories. Given that many scientists would like to understand gauge theory better, it would be totally crazy to abandon String Theory!

How convincing was my story? Well, I'm just a random dude -- another talking head online. What was all that gibberish about SYM and supergravity? Why should I care? These are tough questions for String Theorists to answer because the public wouldn't really understand the answers. They also probably wouldn't be able to distinguish the good answers from crackpot ones.

I wish I knew how to solve this problem. One thing we scientists can do is try to win trust by giving good, understandable answers to questions that do have understandable answers. Once people believe we know what we're talking about, then we can do our best not to abuse that responsibility and to give good advice about questions whose answers are harder to understand.

So go ahead and look through my Quora answers and decide if you can trust me. And if you can, then trust me on this: String Theory lives.