Recognized tokens and data security

Hi,

Would the use of tokens for names (as well as for figures or dates) make it harder for hackers to exploit data in transit?

I see there are ways to program tokens for personal names:

https://community.sdl.com/solutions/language/translationproductivity/f/90/t/5923

I'm interested in doing what I can to make translation memories safer, like removing names, addresses and other identifiers from our translation memories. I've developed a few tricks to find content with personal data that was accidentally translated in a translation memory meant to contain no identifiers, it's a boring but doable job if you only have a few stray confidential documents in a memory that is not too big. 

Am I right to think that using tokens for personal names would help with productivity but also make those names less searchable like tokens seem to do for figures and dates?

Thanks

MP