Hey,
I went to that website and it seems to say that he has stopped working on this. Other places on the internet I found that no new maintainer has been found. Try to google "xesearch unmaintained" and you will find several hits regarding that.
I actually did something similar when writing my master thesis in anthropology about a small village in the USA as I needed to anynomize a number of names in the entire text, so I changed the names, but I wanted to write the original names.
I did this without any packages. I have addd the description of this below.
For lualatex there is however a package that does this -- chickenize. Unfortunately Ubunut doesn't have the current version of that packaged, but if you include the attached file with a simple
\include{substitute} it works the same (I just put this together tonight)
\addtosubstitutions{Fisher}{Hawthorne}
and then once:
\substitutewords
This will replace all instances of "Fisher" with "Hawthorne". I have also attached an example file (letter_to_fisher.tex) to demonstrate how to do this.
Googling further, it seems also Xetex is kind of going towards unmaintained. One of the main reasons I started Fidus writer was precisely that the 1500 ways one can create a PDF file with LyX are just too much for any normal user. For the user who is able to navigate all those and does write latex code and absolutely needs to use an older latex rendering engine, the best is probably too use search and replace on the latex output files of Fidus Wirter and turn them into what one needs them to be.
-----
My thesis way of doing things:
The way I did it was like this:
\def\anonynomize#1#2{\expandafter\def\csname #1\endcsname{#2\xspace}}
\anonynomize{John}{Curtis}
\anonynomize{Fionna}{Maggy}
\anonynomize{Vera}{Michelle}
\anonynomize{Angela}{Maria}
\anonynomize{Garst}{Daniel}
\anonynomize{Julius}{Jonathan}
\anonynomize{Jacob}{Garst}
\anonynomize{Abraham}{Alex}
In the text I could then write
"And there \John came" and in the output it would turn it into "And there Curtis came".
And yes, I realize that is one "\" more than in your current setup. The advantage is that you don't accidentally convert something that wasn't meant to be converted.