My understanding is that part of the issue with string theory is that the computations are extremely difficult. Like, yes, the landscape is very large, but it’s also very difficult to, even picking one particular option, it is very hard to get a prediction out of it. However, I think some correct but very loose postdictions have been made about some mass ratios?

There are many critics of string theory, and their voices deserve to be heard. However, I would also like to ask these critics: Can you come up with something better? String theory explains many phenomena with mathematical elegance, but at the moment, there is no way to test it.

Von Neumann's elephant applies here. ΛCDM only works because they tune hyperparameters. I don't really see string theory as any different - why do we think that space is composed of tiny coiled springs? That's just a random thing we invented mathematically & then made the math work. It does not mean we started from some fundamental understanding about a real property in the universe. The Ptolemaic system is mathematically correct if you let the math get complicated enough and tune enough parameters (in fact it's mathematically isomorphic to the heliocentric model due to relativity), but it doesn't mean it's the right philosophical framework to think about things.

In other words, even if the math for string theory does end up making testable predictions that others can't, it's not necessarily true that string theory isn't isomorphic to some simpler conception of reality.

ΛCDM is not a parameter elephant-fit. Yes, it has six(ish) free parameters, but those parameters were fixed two decades ago by the WMAP+2dF data set. Since then the model has predicted not retro-fitted dozens of independent observables: the precise locations and amplitudes of the CMB acoustic peaks (Planck), the BAO scale in galaxy surveys (BOSS/eBOSS), the lensing convergence power spectrum, the late-time damping tail, even the primordial helium abundance. Every new instrument is effectively a blind test, and ΛCDM keeps passing with the same numbers typed into the input deck. That’s quite the opposite of parameter-cranking.

I’m not claiming it doesn’t have predictive powers. But the theory does parameter fit through the addition of dark matter and dark energy into the model, which we don’t know if they exist since we have absolutely zero theoretical description of what they would be. And 85% of the observable universe is supposed to be this unobservable thing we have no model for? That doesn’t feel like parameter fitting to you?

And while it does have predictive powers, it’s equally important to be honest about its failed predictions which have skyrocketed since Webb conclusively showed the age of the universe doesn’t fit that model. That tells me there’s an elephant we haven’t found yet.

The optimist in me believes that String Theory is the physics of universes, not just our own which is why it has so much fine-tuning needed.