This year I got few cases that Oracle DUL cannot recover the data well for you, while AUL/MyDUL can do it well. In the following cases, you should choose AUL/MyDUL as the recovery tool.
CLOB data in variable-length character set database. Oracle DUL cannot process the character set difference stored in CLOB columns automatically, so you may get the mess code for CLOB values. If you use fixed-length character set, both AUL/MyDUL and Oracle DUL are ok.
NCHAR/NVARCHAR2 data. Usually NCHAR/NVARCHAR2 data is stored in different character set, while sqlldr cannot process two different character set text in the same file, and the Oracle DUL's dump format does not support NCHAR/NVARCHAR2 type, so Oracle DUL does not support it, it will treat them as CHAR/VARCHAR2 type, finally you will get mess data again.
LOB data without system tablespace. AUL/MyDUL can easily borrow table structure information from existing database, and then you can manually specify the table segment and LOB segment information and finally get the LOB data back for you.
However the latest Oracle DUL version support ASM, but I am not sure of whether Oracle DUL support corrupted ASM, if ASM is not corrupted, we can copy the data files out to file system, then AUL/MyDUL can recover it.
And Oracle DUL have "unload schema" and "unload database" command, AUL/MyDUL can only unload table one by one, but AUL/MyDUL have "list" command and run script file command, it's not hard to unload a schema or whole database at all, this feature is not important at all.
AUL/MyDUL does less corrupt block check, so you will get less error message during the recovery, and can get the maximum data back for you. Some Oracle DUL users reported it in the above cases.