Billions of dollars leading to failure during clinical trials is the overwhelmingly likely outcome for all therapeutic areas, particularly so for CNS.
Somebody else asked "what is your alternative?" and you responded with "drunk under a streetlamp". The issue with this is that all of science is drunks under streetlamps. Scientists rely on a set of methods; for preclinical work, this is animal models, organoids/tissue chips (still in development), and in-silico AI stuff. Slim pickings! To use your lamp analogy, all of these options will give off a pretty dim amount of light.
What evidence would it take for you to reconsider how we are approaching Alzheimer's?
Our working theory is that amyloid beta plaques cause Alzheimer's, and so preventing them will prevent the disease. We have absolutely demonstrated our ability to prevent those plaques in both mouse models and humans. Doing this cures symptoms in mouse models. It does not cure symptoms in humans. I already gave a list of 9 different approaches that generated this exact result. How many more billions do we need to spend on new approaches to do the same thing, before the field reconsiders the theory?
And the failure here is not just with failed drug trials. In this century, it is easier to find fraudulent data advancing the hypothesis, than real data. With the result that several fraudsters have managed to create very good careers for themselves before being caught. This is a sign that there is something very deeply wrong with the field, if such fraud is rewarded rather than being caught more promptly.
Furthermore you are absolutely wrong about what I said about the alternative. My drunk under a streetlamp comment was about continuing to use these animal models to create more drugs. My suggested alternative was basic science. If amyloid beta plaques are a symptom, not a cause, of Alzheimer's, then what is the cause? We need to figure that out, before trying to create new drugs for it.
And it isn't like basic research has nothing to try. Alternate hypotheses with some suggestive data for them include Tau folding, inflammation (possibly caused by viruses), problems with the blood-brain barrier, problems with the mitochondria of neurons, dysregulation of metal ions, and so on. Maybe the real answer is not even on this list. But finding and demonstrating the real cause should be a high priority.
My father-in-law has Alzheimer's. If research continues for the next 30 years like it has in the last 30 years, there will still be no treatment when my wife gets old enough to worry about getting it. This is not a potential future that I'm particularly looking forward to. (But maybe I'll get lucky and get it before her...)
The researchers, as quoted in this article, view their work as a step towards understanding the interplay between amyloid proteins, the BBB, and cardiovascular dysfunction. That sure sounds like an interesting step towards deeper root causes.
The fact it has to be done with mice is the best we can do. The article also does not elaborate on what interesting modeling or synthesis techniques might have been used, which could be more generally applicable.
Somebody else asked "what is your alternative?" and you responded with "drunk under a streetlamp". The issue with this is that all of science is drunks under streetlamps. Scientists rely on a set of methods; for preclinical work, this is animal models, organoids/tissue chips (still in development), and in-silico AI stuff. Slim pickings! To use your lamp analogy, all of these options will give off a pretty dim amount of light.