I just got the large chests, and I'm very disappointed. The 2x2 chest only holds 40 stacks, the same as two iron chests. The 3x3 holds 60 stacks, the same as three iron chests.

But my most common use for chests is to put a row of them on the side of a machine and/or train for buffering purposes, so this makes them a pure downgrade from iron chests due to requiring more space for no extra storage. In fact, with the 3x3 chest I could even have doubled up on iron chests to get twice as much storage in the same space! (Yes, you need more inserters, but those are cheap and easy to spam.) That's on top of the higher tech and material cost, which makes the large chests something of a noob trap.

The only possible use cases I can see for the large chests as-is would be for being able to easily load and unload at a ninety degree angle, and possibly for use with loaders? But I don't use those (I prefer direct train->chest->machine->chest->train buffering) so they are all but useless to me in their current form.

I understand that chest volumes are kept low deliberately in this mod, but the big chests need to have more in order to not be a waste of space and resources. The bare minimum I see for a possibility of usefulness would be 50 stacks for 2x2 and 150 for 3x3, giving the big chests a little more storage space for the same machine "frontage" space used compared to iron chests (and 3x3 comparatively more than 2x2).

Although, that would still be less storage space per volume (including a row of inserters on opposing sides) in the case of 2x2, which would be rectified if you gave it 60 stacks - which makes it more dense than 3x3 in terms of just the chest area, (but not if you also include the inserters). And it's still less storage per frontage than even a double-row of iron chests. I feel like that puts it in a good spot to have some consideration and tradeoffs with both iron chests and 3x3 chests.

If you feel like 150 is too high for 3x3, it still needs to be at least 120 to match a double row of iron chests; although at that point it's only saving on extra inserters. A modest but extant usefulness, I suppose.

So, yeah. I strongly request you increase the 2x2 chest volume to 50-60 stacks, and 3x3 chest volume to 120-150 stacks. (Personally, I prefer the higher ends of the two.)

The fact that you can fit 12 input/outputs around a 3x3 chest is way more useful than how much they can actually hold.

I don't see how that's all useful - or even true in a practical setting.

If you're feeding a machine from chests, the feed rate depends only on the length of the side which faces the machine. You can get the exact same effect from a row of iron chests, for less overall footprint - and given that you probably want to have maximum inserters possible for max feed rate in this mod, there's no downside on that front.

If you're feeding multiple machines, why can't you use multiple iron chests, or better yet a belt? If a belt is not high enough feed rate, then I question why you're trying to fit that many machines around a single chest. The only thing that must be shared between machines is the cube, after all. Or why not try to feed the machines from a train, instead? That's what I'm doing.

Speaking of, I have actually found the added space of a 2x2 chest is useful for padding the odd-width machine's footprint to perfectly fit between two parallel rails (as the gap must be an even number of tiles). A purely coincidental usefulness of the 2x2 chests... Even if I could do the same with belts, it at least looks nicer.

A chest larger than 1x1 is effectively an auto-sorting, configurable item teleportation machine that can handle many different types of items in parallel. It's a very powerful design tool even in standard factorio. See warehouse malls e.g. https://www.reddit.com/r/factorio/comments/1408zpl/se_warehouse_mall. If you could have a single chest as large as you liked, there would be no need for any other logistics at all. It's intentional they're less space-efficient as a tradeoff.

Also, I didn't want to particularly encourage players to make huge storage buffers by providing massive chest inventories, since while some amount of buffering in ultracube is generally desirable, going too crazy with it could lead to a factory that's sluggish to respond to changes and harder to assess.

If your goal is to pack the most items you can into one tile, the iron chests are there for you :)

I can see that with a single 7x7 or larger chest and multiple 3x3 assemblers, but we're talking about the complete opposite situation here. The most important machines, the ones driven by the cube, are all 7x7, and only one can operate at a time. And even the Fabricators are 4x4, so it's impossible to get more than 4 around a 3x3 chest—so the only advantage there is not needing bots to supply it.

As you noted, buffering is very important in Ultracube, although you don't want to go overboard with chest space density for multiple reasons. The former is why I expected a practical increase in storage density with chests that cost much more in terms of research and resources, and felt cheated by a "noob trap" when the investment I made to get them before other important things like robots proved to be wasted. After all, no matter what tricks you can pull off with large chest footprints, their most obvious and primary function is storing things—so why shouldn't the more expensive chests be better at storing things?

You don't have to go overboard with it; I believe a balance can be struck. If you don't want them to offer a better mass storage solution than iron chests, then 60 and120 stacks would make the larger chests just on-par with the iron chests, and I can prove it with math:

Consider the setup of rows of iron chests, with rows of inserters between them: this arrangement is probably the best way to store as many items as possible in minimum space, and is the use-case we want to avoid making the larger chests better at (and thus encouraging the employment of). Now, with many rows of iron chests, the ratio of chests to inserters approaches 1:1, and therefore we can approximate the density as two tiles per chest. Since each iron chest stores 20 stacks, that means the theoretical limit of bulk storage density for iron chests is 10 stacks per tile.

Now, let's consider larger chests in the same configuration of rows of chests linked by rows of inserters. There may be gaps in the rows of inserters, as only one is needed to link two chests, but there is no easy way to make use of that space for more storage. Therefore, if we apply the same process as before of looking at the limit of the ratio of chest-tiles to non-chest-tiles as the number of rows increases, we get 4:2 for 2x2 chests and 9:3 for 3x3 chests, getting a net 6 tiles per 2x2 chest and net 12 tiles per 3x3 chest.

Therefore, if the 2x2 chests could hold 60 stacks, and the 3x3 chests could hold 120 stacks, they would only achieve the same density limit of 10 stacks per tile. Therefore, they would not have greater utility for the bulk-storage situation you wish to discourage.

However, the greater stack size per chest would make the large chests more useful in the use-case you wish to encourage: small buffers of only a few chests. Especially ones that aren't attempting to maximize storage density and therefore are more likely to have access to all four sides of the chest.

Therefore, I contend that raising the stack sizes to 60 and 120 will be a net positive for the mod.

there is no easy way to make use of that space for more storage

Disagree. If you place chests with offset, you can fill that space. Although, it will have less convenient (un)loading setup.

For 2x2 example with items movement left-to-right:

1) place chest A

2) place chest B 3 tiles to the right and 1 tile down from A

3) place inserter between them

4) place chect C 1 tile to the right and 2 tiles down from A

5) repeat steps 2 and 3 relative to chest C instead of A

Now you can tile this setup in both directions (horisontal and vertical).

Therefore the net limit is 5 tiles for 2x2 and 10 tiles for 3x3 chests.

So 50 and 100 stacks, perhaps? Although personally I don't see the extra 20% over theoretical as too drastic, especially considering the higher cost of the larger chests.

Sorry for not reading this earlier. You argued your point very well. I'll increase the capacity so that the bigger chests at least match the stack-density-per-tile of the iron chests with the obvious setup.