how do you delete one or more letters from a panda series?

You can use \b for word boundary, so 'xxx abcd' can get replaced. Since the x and can be interleaved and you would remove them all, you can just use:

s.str.replace(r'\b[x\s]+\b',' ')

Output:

0    transworld system inc trying colect debt mine ...
1                  complaint cals ocur betwen cel job 
2                                     company violated
3    previously requested experian actualy mine ins...
4                         company trans union noticed 
dtype: object