I ran out of time myself. In my opinion, this was a great game choice with great technical documentation, but poor casual play documentation for the allocated amount of time.
After getting familiar with the documentation and basic playtesting for 30 minutes, I figured I would never win going with the wrong warp + sniper rifle through wall strategy since I don't have much experience in TASing in general. However, since the source code and symbols were provided along with one known ACE strategy, I tried doing that instead given my skills.
My plan was to execute 3 bytes of code using the torches room, since it seemed well documented and that other methods I could thin of didn't seem reliable enough. I decided to go with STA 0x0322 and hope that the accumulator happened to have a non-zero value by luck, and if that didn't work hope to have enough time to hex-edit my input file to try something else. Since this is 8D2230 in byte code, it was just enough bytes to encode using the torches. After figuring out which torch corresponded to which bit in memory (the 8 left torches are stored in 0x0302, middle in 0x0303 and rightmost in 0x304, with the least significant bit of each group in the top-left corner, increasing from top-to-bottom, left-to-right), I ended up with this torch pattern:
10 00 01
00 11 01
10 00 00
11 00 00
Note that I also considered executing STA 0x0300 since it was faster input, but I believe it would have lost time for reaching the winning condition, since the game would have needed to wait for the dying boss counter to decrease enough times.
On a side note, while researching the torches room, I stumbled on a page that wasn't included in the documentation and that explained
the solution to the torches puzzle with a video. (It's returning an HTTP 403 Forbidden error right now, not sure why.)
Figuring out and planning all of the above took me 2.5 hours. That left me with 1.5 hours to plan my route from the start to the torches room, then to a shop to execute the inventory overflow glitch, and go back to the first screen to trigger ACE. This is when I realized that money would be a bigger issue than I expected due to the price of the SMG. I then took 30 minutes to find and analyze the drop tables and figure out which enemies dropped what. I figured that zombies are good to farm early on for item variety and crafting, while spiders are good for sell-able gems.
I then started my initial route to the torches room and towards the closest shop. Including decoding the map using mapcreator.html, getting familiar with how to optimize input decently and doing so, this took me an additional 30 minutes. This is when I realized that I had much fewer drops than I expected. I re-checked the drop tables only to realize that there aren't enough item drop types to fill my inventory, even after crafting all items, buying all the unique items from the 2 easily-accessible shops and using the gun duplication glitch twice, and that I lacking much more money than I thought.
With only 30 minutes left, all I could do at this point was going in a wild-goose chase around the map hoping that I would stumble on some amazing loot at a scripted location, since I didn't have enough time left to analyze the code. I failed horribly in doing so since all the chests I found had nothing of interest. Only then did I learn about the time extension, but even with the additional 45 minutes I still couldn't figure out where any of the good stuff is in the game or extract that information from the source code. I used the remaining time to check if I could figure out some kind of money underflow glitch and verify if the Konami code was disabled correctly without success.
In the end, I wasn't able to get enough resources, I still don't know where these resources are (including the sniper rifle and any of the keys), I didn't had the time to consider analyzing the RNG even though it was simplified for the competition, I'm still don't know if my strategy would have worked since I could not test it and since the item to use for the overflow glitch wasn't documented, and I feel that I would have needed all the allocated time just for a casual playthrough and even more to document my findings. My impression is that almost all entrants will have picked the sniper rifle + wrong warp route, while a handful of remaining entrants will be teams composed of a strong reverse-engineer and at least another person that were able to find a good ACE route in time.
My conclusion to all this is that while it still was fun for me most of the time, it would have been a much better competition if the following would have occurred:
a) A complete strategy guide for casual play would have been provided.
b) The competition would be in split categories or limited to a single category.
I'm looking forward to the final results. My bet is that the human players will win if they don't make a major mistake due to their greater knowledge of the game.
EDIT: Whoa, I can't believe I already missed it while writing this post!! I will need to watch the VOD.