I have something that actually kind of works, but you will need to add another Evaluator Column to do this.
The Evaluator Column stempart returns the filename stem minus the number(s) and everything after that if any.
So in your example it would return this:
Here's the demo column.
The column doesn't have to be in the file display for this to work.
<?xml version="1.0"?>
<evalcolumn align="0" attrrefresh="no" autorefresh="no" customgrouping="no" foldertype="all" header="stempart" keyword="stempart" maxstars="5" namerefresh="no" reversesort="no" title="stempart" type="0">n = name;
x =RegEx(n, "(.*)( [0-9].)", "\1");
y = RegEx(x, "(.*)\..*", "\1");
return(y);</evalcolumn>
And then this is the duplicate filter:
Edit: Another approach would be to isolate the words in filenames, but that would require scripting. There is a thread kind of related, but far from this, that attempts to select files with words contained in a selected file. I still have it because the results sometimes make me laugh.