Alas, when the set was finally released and we got to see it for ourselves it quickly became painfully obvious that it was, to put it mildly, not what we had hoped for. Even leaving aside the production and editing issues that are endemic to everything Troll Lord Games has ever published (among others things, the adventure was double-statted for both the then-current D&D 3rd Edition and Gary's own Lejendary Adventures game, which was handled very awkwardly and made things not only much longer but also much harder to read and comprehend) the content of the adventure was sorely lacking.
There are 49 panes, 7 colors and 7 shapes in every combination, and to finish the adventure the PCs must successfully complete all of them. Each plane takes the party to a discrete sub-plane, usually dropped into the middle of a scene where they first have to figure out what's happening and then what they need to do to escape (usually performing some task or defeating some enemy). Many of the panes put the PCs into other bodies - animals, monsters, other humans, and in one case even animated acorns. Some of them require solving puzzles of various types. Some are straight combat. So far, so good. But on closer examination some problems become apparent. First and most obvious is that there are way too many panes, and the requirement to complete every one of them shows that the "significant playing length" description wasn't kidding - actually playing through this adventure would likely take two years or longer. Second, too many of the pane adventures are too similar to each other, which is likely to become tedious. Thirdly, the entire thing is completely arbitrary: the situations are all random and unconnected, with no larger scale meaning or pattern, not that it matters since they have to complete all of them anyway. Fourthly is that many of the pane-adventures are significantly underdeveloped, more sketched and fully-written, with a requirement that the individual DM either add a significant amount of their own development or treat things in a very shallow and railroad manner - and the fact that every one of the situations must be played through in order to finish the adventure makes this a pertinent issue, because the DM can't just choose to skip over the weak or problematic entries. [There's also a technical issue in the D&D rules context in that, over the course of playing through 49 panes (actually 51, including two special panes that appear when certain conditions are met), it's inevitable that they'll gain several levels; since the panes may be entered in any order the party will likely be several levels higher for the later ones than for the earlier ones, which the adventure glosses over in what I consider to be a very unsatisfactory manner by simply recommending that the DM increase the difficulty of the encounters within the panes to keep pace with the PC levels. I can only assume that in Lejendary Adventure, the native ruleset under which Gary originally wrote and played this material, character improvement is on a shallower curve and this wasn't an issue.]
Because of all of these issues it was obvious that I would never actually run this adventure, so it went off into the closet and was pretty much forgotten about. Which was unfortunate, not only because it so badly failed to live up to my pre-publication expectations, but because despite the overall disappointment of the product as a whole there were a few episodes that did capture moments of the old Gary magic. Those feelings were exacerbated when Gary died a few years later and this was left as his last major legacy (since the later Castle Zagyg series was both left incomplete and was in large part the work of Gary's co-writer). So, from time to time over the years since I've pulled this out again, or at least thought about it, convinced that this could have been better.
Eventually, I sat down with the adventure for a couple of hours and pulled together the following, my ideas and notes for how to extract the actual Gygaxian classic buried within the mess that was published.
My first recommend change is to reduce the number of panes significantly. Instead of 49, I've settled on 25. My second change is to tie the panes together thematically by both color and shape, as follows:
SHAPE (Challenge Type)
Rectangle - combat
Square - helping
Oval - puzzle
Disc - transformation
Hex - misc
Star - false finale
Diamond - finale
COLOR (Setting)
Red - underground
Orange - misc
Yellow - desert/plains
Green - woodland
Blue - mountains/boreal
White - false finale
Violet - finale
Combining the two, I came up with the following "best of" panes list.
Red Rectangle - Pane 35 (Key)
Orange Rectangle - Pane 30
Yellow Rectangle - Pane 19
Green Rectangle - Pane 3
Blue Rectangle - Pane 11
Red Square - Pane 47
Orange Square - Pane 27 (Key)
Yellow Square - Pane 44
Green Square - Pane 17
Blue Square - Pane 4
Red Oval - Pane 20
Orange Oval - Pane 48
Yellow Oval - Pane 21 (Key)
Green Oval - Pane 5
Blue Oval - Pane 16
Red Disc - Pane 10
Orange Disc - Pane 46
Yellow Disc - Pane 22
Green Disc - Pane 1 (Key)
Blue Disc - Pane 13
Red Hex - Pane 8
Orange Hex - Pane 15
Yellow Hex - Pane 45
Green Hex - Pane 36
Blue Hex - Pane 23 (Key)
Violet Diamond - Pane 51 (appears when 5 key panes completed)
The "key" panes (one of each color + shape) become the only mandatory ones needed to activate the final pane and complete the adventure, so a group of players who are paying attention can figure out and exploit that pattern. In order to draw further attention to the key panes and emphasize the pattern of them, I also recommend that the first four completed key panes also provide the party with special useful tokens:
Special Tokens (gained upon completing first four key panes; one use each)
1. Lens - view a pane before entering
2. Reset - return to pipe without completing a pane
3. Extra Life - killed in pane respawns in pipe
4. Phone a Friend - commune with Zagyg
This doesn't solve all of the issues with the adventure - not all of the selected panes neatly fit the challenge or setting categories I've placed them into, and many of the pane-adventures are still underdeveloped (and of course to actually run this I'd also need to convert all of the stats into a game system I know and use) - but I feel that it solves the core problems, and transforms the unusable mess of the published version into something that I could use in a game, an adventure that would be challenging and memorable to the players without overstaying its welcome, and that would actually feel like it was rewarding player strategic planning and attentiveness and not that they were just passengers on an inevitable train ride. Something like this would be, I feel, a much more fitting legacy to Gary Gygax's memory than the unfortunate product that Troll Lord Games actually released.
P.S. Lastly, the backstory of the adventure - the trapped deity, the reason why he was trapped, and why the PCs are sent to rescue him, involves the machinations of various deities. In the published adventure, these deities all come from Gary's Lejendary Earth pantheons. If the adventure is instead adapted to the World of Greyhawk, I recommend the follow deity substitutions:
Deities Gwynn = Baalzebul
Bili = Belial
Arianrod = Fharlanghn
Llew Llaw Gyffes = Pelor
Amadan Mors = Zagyg
McGreggtim = Heward
