Argh - I just wrote a response the size of a tome and deleted it.
In short, Wiley discusses this issue at length. His evidence generally favors the traditional attribution of work, although it seems that Petipa had outlined the lakeside scenes and even contributed written descriptions of certain choreography, including the entrance of the swans.
A number of Ivanov's ballets were notated, particular the one-acts. His larger works like TULIP OF HAARLEM appear to be lost. Some didn't continue in rep and some have been preserved simply because they were in rep during the notation period.
Wiley has made a couple of attempts to compare Petipa and Ivanov and has used Petipa's Shades compared to Ivanov's Snowflakes. In my opinion, this is like comparing apples to oranges. A better comparison might involve Petipa's waltz from LE JARDIN ANIME and Ivanov's Snowflakes, which are both large waltzes using minimal properties (no stools, but either garlands or wands). The SLEEPING BEAUTY waltz could also be used but it is less "dancy" than the other two. This comparison reveals a choreographic similarity, but possibly the similarity is a result of the normal choreographic response of the time to music of the same genre (a waltz).
All in all, Ivanov was the assistant, Petipa was more famous and was doing more work. He was the dominant figure at the Maryinsky. Petipa also appears to have been much more driven than Ivanov, who liked to play cards during rehearsal (even rehearsals he was running

). If Petipa felt the newspaper had printed an incorrect attribution of work, he would write to the editor and say so.
SWAN LAKE wasn't the big deal in its time that it has since become. I think the contributions of the choreographers have been magnified and more importance has been attached to Ivanov's reputation based on SWAN LAKE than is warranted in view of his entire ouevre. I'm not saying this is a bad thing, but simply that SWAN LAKE is/was not the make-or-break ballet of Ivanov's career.
[ 07-02-2001: Message edited by: doug ]