Several bugs were fixed in the CSV and TSV formats and the performance was dramatically improved between version 10 and 12. Leading and trailing spaces that are part of non-quoted fields are ignored as this is by far the most common behavior and expected by many applications. The old implementation of CSV was based on the Table format, so you can get that behavior bug for bug by doing:
Thank you both for the quick responses with suggested solutions and explanations.
I will then adjust my evaluations and pay attention to the speeds. The data sets are many MB in size and consist of decimal numbers, hexadecimal numbers and strings and must be converted accordingly.