What about a database of people from countries with high propensity for radicalization, for example Tunisia, or Somalia?
Paradox: Tunisia is perhaps the most progressive of all the M/E Muslim countries, and yet they are the #1 M/E country in terms of Jihadis. They are overwhelmingly the greatest per/capita Jihadi producers in the world. By far. Tunisia is very small. Networks are very small. 95% chance that a young male Tunisian knows someone off to fight Jihad. This is very significant.
Somalia - is a stateless war-zone, people coming from their are massively over-represented in other kinds of crimes. Wherever there is a 'new' Somalian community, the crime (drugs and related) offenses are off the charts with over-representation. In Toronto - a big slice of the 'violent murder' pie is committed by Somalis - and they are near 0% of the actual population. Some social workers have observed that they simply have absolutely no respect for authority - it's meaningless to many of them. There is an overrepresentation among Somalis in terms of 'domestic' (i.e. host nation) violence, that's possibly 'terror' related.
Those are both good examples because it shows you how young men from completely different places, are radicalized for completely different reasons, and are associated with completely different kinds of violent people - and their nationality (read: culture) is definitely a correlating (or even predictive factor).
It's a moot point anyhow. The FBI has a database. You're probably in it. There's a tag on your file that might indicate the likelihood you'll be radicalized.
We have to be really quite careful of this kind of stuff, frankly, I'm more concerned about the 'populist anti-muslim sentiment' that might arise, as opposed to any kind of actual 'database' of muslims.
Anyhow - it's all moot. There will be no DB of Muslims.