I am not too keen about 6.0-0, but in any case Black has a fine game after 6.ed5 Qxd5 7.d4 (7.0-0 e4!) Bg4! for example 8.c4 (8.Bc4 Qe4+ 9.Be3 Bxf3 10.Nd2 Bd1!) Qe4+ 9.Be3 Bxf3 10.Nd2 Qg6 and Black has some advantage. These are rather old analyses, which are repeated in the recent repertoire book by Kaufmann.
Harding's books are rich in content, but full of analytical mistakes- I'd rather say they do not stand up against silicon aided analysis.
I have played the Ponziani for years, and I actually own one of the few books ever published on the opening by TD Harding. During an online tournament the following position arose from 1.e4 e5 2.Nf3 Nc6 3.c3 d5 4.Qa4 f6 5.Bb5 Ne7 6.exd5:
After the game I asked for a computer analysis, which I always do because I'm naturally curious about where I may have gone wrong or where I could have improved. Interestingly, the computer noted 6. exd5 as a mistake, and that 6. O-O, was stronger.
I thought that was interesting because I have an old, ratty copy of TD Harding's book that lists exd5 as being among the strongest replies. In fact, when I decided to search the database for this position, there were no less than 51 master games in which white plays 6. exd5. In those games, white scored 47.1%, drew 17.6%, and lost 35.3%.
Which leads me to this question...if the move is as solid as all the research suggests, why is the computer labeling this as a mistake?