Since you've used my code (which you can see works in the image I posted) and you say referencing "tFileList_1_CURRENT_FILE" works after tAggregateRow, the probable reason you're not getting the file name is that you haven't included FileName in the Group By part of tAggregateRow.

PS: Glad to help; it was a far more interesting question than the usual for this forum

Using the following test files, where the first row of each file is assumed to be the header:

test.txt
A|B|C
1||3
|2|3
1|2|3

test2.txt
F|A|X
|A|B
C|D|E
F|G|

the following job does what you want:
- tFileInputFullRow uses (String)globalMap.get("tFileList_1_CURRENT_FILEPATH") for the file name
- row2 schema is FileName and line
- tNormalize normalises the line column separating on "~~"
- tExtractDelimitedFields splits the line column separating on "::"
- row4 schema is FileName, ColName and ColValue
- tReplace searches for whole word "" in ColValue and replaces with null
- tAggregateRow groups on FileName and ColName and outputs count of ColValue ignoring nulls
- tJavaFlex code is as follows:

Start code
-----------------------------------
int RowNum = 0;
String[] ColNames = {""};
int ColNum = 0;

Main code
-----------------------------------
RowNum++;
if (RowNum==1) {
    ColNames = row1.line.split("\\|",0);
    row2.line = ColNames[0] + "::"; }
else {
    ColNum = 0;
    row1.line = ColNames[0] + "::" + row1.line;
    Integer DelimIndex = row1.line.indexOf('|');
    while (DelimIndex>=0) {
        ColNum++;
        row1.line = row1.line.replaceFirst("\\|", "~~" + ColNames[ColNum] + "::");
        DelimIndex = row1.line.indexOf('|'); }
    row2.line = row1.line; }
row2.FileName = (String)globalMap.get("tFileList_1_CURRENT_FILE");

Just replace tLogRow with tOracleOutput (assuming your table has the ID field set to auto-increment) and voilà.