I don't believe this was done intentionally, but instead a side-effect due to the way structures work. After it was discovered, a group effort was made to surround the perceived unraidable base West of spawn SPECIFICALLY to deny Teleportation inside of the base. It wasn't until later, that it was discovered that all permissions were overwritten other than the ability to break Iron Blocks. My team fell victim to this, as we were unaware that blocks other than Iron could be broken. Since then, our structures have been made with 2 layers thick of Iron Blocks to prevent further abuse of this mechanic. I think if this were to be changed in the future, that only the "enter" permission should be overwritten. Meaning players making structures around other structures can only prevent other players from entering a structure. This way it can't be abused to obtain loot, but can be used as a strategy to deny warps or storage. Players would be encouraged to have warps around their structures if they are "surrounded" and you are not able to TP directly inside of them. Attackers could make measly -75% structures that will inevitably be taken down to temporarily setup ganks for example, but in the actual case that happened on the server, it took a combined effort of over 31 STACKS of Iron Blocks to "block out" the base West of spawn. If players build their structures an optimal size for +100% while still being big enough to not easily be surrounded, I think it adds an extra layer to defending a structure. That's how I feel, but if it came down to all or no permissions overwritten I would be for changing the current "overide" to not overrule any permissions.
The theory you suggested would technically work, but keep in mind the "alternates" still need to own less than 50% and with your suggested method would have diminishing returns. This is because when they invest again you can own just under 50%, but you already invested Rupees previously which counts towards this total. I tested the theory, and it works, but each time Alt #1 and Alt #2 invest they have to put in less and less. The total Structure value is increasing, but the ratio that the other accounts can own compared to what they already put in still plays a factor. With large scale investments you can see the diminish more than when you test it on a small scale. However what you said about the structure essentially having a +0% is correct for 3rd party investors, but keep in mind outside investor's Rupees are always supposed to be counted as pure. Your method works, but using more accounts opposed to carefully calculating between 2 is vastly more effective when protecting very large -75% structures. I understand some players might not have access to an entire team, but even with your method, the original investor is still losing 75% of the rupees they put in. When speaking about actually wanting to protect bases and the original investor puts in 100k for example, you're still losing that 75,000. I would still consider this somewhat of an "exploit" due to being able to get more out of your Rupees if you have multiple people helping, but taking over structures were designed this way so it is more of a tactic for those that know how to pull it off. Due to how structures are taken over as of now I can't imagine anything changing, rather this being something to be aware of.
As much of a cool idea this would be for making rigged bases, I'm not sure how I would feel about it if it were actually implemented. First, I don't think explosion damage for players should be able to be toggled on and off. A player should be punished for not killing a Creeper, accidentally setting off TNT, or allowing their base to be TNT cannon'd into. Currently there are only 4 very general permissions and adding 2 just for explosions seems unnecessary. However if toggling Block Damage were to be implemented, I'm not sure if traps would be as effective as you'd like them to be. If surrounding another structure and overwriting permissions isn't changed, you could make a bigger structure, turn off explosion damage and then continue a raid, disabling the trap entirely. If that mechanic is fixed, raiders could use Creepers or bring their own TNT to try blowing up the walls of a structure beforehand, testing for explosion damage being on and if it is a trap or not. If explosion damage is able to be turned on, it should also effect all blocks, including Iron Blocks. Turning on explosion damage for a structure but not having it effect Iron Blocks seems like an imbalance and it would be extremely easy to see a trap, especially if the floor or walls were different material. If a player is setting a trap, everything should be blown up, including their investment of Iron Blocks, so it's not super easy to reset an effective trap. This would force players to make a sacrifice other than just TNT and decoration blocks when making a trap. I think it would be interesting, but easily countered by experienced players who would know how to look for traps. Even with what I said considered, I'd still be up for this because who doesn't want to KABOOM scrubs that don't care to learn the Meta.
I think you covered this pretty well, but when this was first discovered, there was a couple more details about what actually Breaks / Reapplies the AI or not. It's also worth noting that switching from the Rotten Zombie Skull to any other Artifact fixes the AI, even if Rotten Zombie Skull is reequipped.
For those who were familiar with the Ender Pearl Floor Teleport Glitch that was popular in many faction / raid servers it is also possible to re-create this effect with BAMF. 1.9+ Ender Pearls have changed to add a cooldown removing the old glitch that has been abused since it's implementation. However, BAMF can still be used with spammable projectiles such as Eggs / Snowballs to achieve the old effect of teleporting through a floor. This can also be done with an Arrow, but requires very specific angles and timing.
As much as I loved this mechanic in MCMMO servers, I'm not sure how I would feel about it fitting into MCRPG. I'm a big proponent of players having to work for their gear and although this would be an extremely useful mechanic, it would take away from players having to work even more to get the edge on other players. I could see this being very useful for getting materials after a raid when a base might be filled with unwanted armor or when enchanting and you no longer want the gear for example. It would be an easy way to recycle materials, which is really good. Perhaps this idea could be implemented into an artifact of some sort that would also be required?
I like the idea of Rupees having more uses, but for the same reasons I mentioned above, I actually think it is better that players need to spend a lot of time / work to achieve the underground farms you mentioned. With the possible addition of Silk Touch Spawners in a future artifact, players would already have the ability to easily make farms / grinders. Being able to use eggs to also move passive mobs easily I think would take away from the effort of players having to spend time manually growing grass or making a Silk Touch shovel.
Currently if players wanted multiple stacks of the same type of mob, all they would have to do is start with 5+ in two different locations. This can be difficult to do, but makes it worth while for players that actually do it. Being able to split a stack of 100 for example could exponentially increase the mobs breeding at one time if more and more groups were made and all bred at once. After a stack of mobs reaches multiple thousands and the breed delay is down to 1 second, players shouldn't really need multiple stacks regardless if loot is the main priority. I do agree with your point about Chickens producing more eggs. However, if X Chickens in a stack produced X eggs as if they were separated could be bad if you had a stack of 10,000 Chickens for obvious reasons. I'm not sure what a good system for this would be, but it is definitely true that a stack produces eggs way slower than ungrouped chickens. At the moment a player can only have 4 Chickens in close proximity producing eggs at once, which can be slow.
Having used Voxel and understanding the limitations of WorldEdit, I fully support this. It has been mentioned in other threads before that it could be given to trusted players, which I think would be effective. I'm not sure what permissions system the server is using, if any, but it should also be possible to limit features within Voxel for the ones that could accidentally cause serious damage.
I've used plugins like this before and it is really amazing the amount of different designs people have for heads and how they can be used for making creations. Even though I have used them before and they would allow builders an extra element of originality, I think it would be better not having them. I only say this because I am personally a huge fan of default looking Minecraft and I think it's always more exciting when players come up with design ideas using the default Blocks / Mechanics naturally provided. However, hundreds of different Mob Heads can creatively be used in so many ways, so I think the majority of the players on the build server should be able to decide whether they want them or not. Although I've never been on the build server myself, so I feel like my vote probably should not even count.
tl;dr: As a whole I think it was a very informal post, good job. We'll probably be making more of these in the future as there are still many, many more less significant bugs and missing information on the server / website that should be addressed. Also those OP banners though.