The issue with the regular expressions you provided is that they are looking for a specific format of the string where the numbers are 4 digits and are separated by exactly one space.

In the first statement, the regular expression '^[0-9]{4} [[:space:]] [0-9]{4}$' is looking for a string that starts with 4 digits, followed by exactly one space, then again followed by 4 digits and the string should end there. So it's not matching any of the strings.

In the second and third statement, the regular expression '[0-9]{4} [[:space:]] [0-9]{4}' and '[0-9]+ [[:space:]] [0-9]+' is looking for a string that contains 4 digits, followed by exactly one space, then again followed by 4 digits.

In the fourth statement, the regular expression '[0-9]+ [[:space:]] [0-9]+' is looking for a string that contains one or more digits, followed by exactly one space, then again followed by one or more digits.

In all cases, the regular expressions are looking for a specific format of the string and not considering the rest of the string.

Here's a regular expression that can help you match any number separated by one or more whitespaces in a string:

\b\d+\b\s+\b\d+\b

This pattern uses the word boundaries (\b) to ensure that the digits matched are not part of a larger number or word and the \s+ will match one or more whitespace.

You can also use this pattern to match any number separated by any non-digit characters

\b\d+\b[^\d]+\b\d+\b

This pattern will match number(s) separated by one or more non-digit characters.

You can test these patterns in your SELECT statement like this:

SELECT 'Employee 346 retrad: 0630 3062' REGEXP '\b\d+\b\s+\b\d+\b'

Please keep in mind that the specific regular expression pattern that you use may depend on the specific requirements of your application. The examples provided are just a starting point and may need to be adjusted or expanded based on your specific use case.