It's just REGEXP ran through PHP (and JS if using browser validation) REGEXP functions. See the below regarding unicode in REGEXP. We have no specific limitation in place preventing this. The limitation is with PHP/JS/REGEXP.

www.regular-expressions.info/unicode.html

Don't understand what you mean. Validation suggestions? You mean the validation errors? If you mean the validation errors they are language strings so you can translate them to whatever you like. If you mean the dropdown that provides a few regexp examples I don't see a reason to extend it any further as we provide a means of providing your own regexp.

To give you an example in your current regexp you're using [a-zA-Z], which is literally matching the characters a-z in their lower and uppercase forms. You can simplify this with a word character match using just \w, which will match unicode letters, numbers, and underscores. If you truly just want a-z, but also want it unicode then you probably need to use [\p{L}].

The REGEXP to match unicode letters is [\p{L}]. This works in PHP REGEXP. It will not work in JS. This means it will not work in browser side validation. For JS unicode I believe you have to specify every unicode character you want to match using it's hexadecimal value. Again, none of this is a limitation of CB. You are only limited by what PHP and JS support.