This question might be better suited to a subreddit like /r/politicalscience. I'll try and give you an answer here, but my answer discusses the past 10 years of the Fragile States Index, and so will probably fall afoul of the 20 years rule.

If you look at the Faq page for the index, the authors plainly state:

It is important to note that these ratings do not necessarily forecast when states may experience violence or collapse. Rather, they measure vulnerability to collapse or conflict.

So, by their own admission the list isn't meant to be predictive.

It is good that they don't make such claims, because in the 11 years that the Failed States Index/Fragile States Index has been published, there has been a lot of unrest that they didn't see coming.

For instance, in the edition published in June of 2010, Tunisia and Libya are given rankings comparable to Brazil. Egypt and Syria are ranked as less stable, but both were improving from 2009. Based on the Index, I don't think many people would have predicted the Arab Spring would happen 6 months later.

Similarly, in the 2012 Index Mali is ranked "79th least stable", marking a slight improvement over 2011, and comparable to Russia or China. This index did not and could not predict that a Tuareg separatist revolt would rise up in the north of Mali, get co-opted by Islamist militants, and provoke a military coup due to the civilian government's perceived ineffectiveness in fighting the Islamists.

Turning to the 2016 Index, Turkey is shown as a middle-of-the-pack country comparable to Peru or Indonesia. That ranking would not lead many to predict that the Army would attempt a coup against President Erdogan.

On the other hand, other countries, particularly African countries, are consistently placed in the "alert" or "near collapse" categories, and yet significant conflict or unrest does not materialize.

For instance, In the wikipedia article you share, the 2016 list puts countries like Cote D'ivoire, Cameroon and Uganda as slightly less stable than Ethiopia or Libya. Cote D'ivoire has had some difficulties with soldiers mutinying due to late payments, and Cameroon has a movement in the English speaking north-west demanding greater recognition from the Francophone majority. But, in my opinion, those difficulties aren't more troubling than an active civil war that is going on in Libya. Nor are they quite as likely to cause revolution, IMO, as large-scale protests among Oromo and Amhara minorities that together constitute about 52% of the Ethiopian population.

So my point is, in a nutshell, that the Fragile State Index is not good at predicting revolutions or unrest now, and I don't expect better predictive ability in the hypothetical scenario where it was developed and applied in the decade before those events.