Allow the user to set a default field separator they prefer to use when exporting CSV files and have Jira observe it.
When the CSV file format was developed to represent spreadsheet data, it was done so with the expectation that most spreadsheets would only contain numbers and short text labels, and since numbers would often contain the decimal point (.) but not commas, the comma (,) would be a very good delimiter.
This is not always appropriate and it would help if the users had a choice of what delimeter JIRA will use.
When opening CSV files in Excel, it looks for a field separator character in the top of an incoming CSV file. If it's there, that's what it uses. If it isn't there, it uses the system default. This is done using a line such as sep=, or sep=;
Rather than inputting this manually, it would be great if JIRA asks the user prior to the export and appends this to the file.