Finally, no objectives are necessary to complete the game, so the players may skip them if there is a significant problem completing it, but it is recommended to at least try because... well, otherwise the game would be boring and mostly pointless.
I'd be wary of basing large portions of the game on "players will do this because, well, it's boring otherwise." It implies that there's no practical reason to accomplish anything and assumes players won't invent preferred goals of their own, like staying alive or killing everything. A more formal system for being forced to abandon major objectives might be better.
But consider--unless the players in any game are being railroaded by someone or the game, they can just choose to go off and do nothing in any game at any time, can't they? Yet, they chose to follow the main quest and often side quests the GM or game developers makes up for them in some way because it's often both fun in a way and it progresses the game. And people play games because they're fun in some way (and those who don't are frankly those I don't care about at all!). I mean, heck, I've watched many SS13 videos, and there's no objectives that MUST be completed as far as I know of and people still play it. Really, I'd even imagine SS13 to be a big sub-conscious inspiration for my idea.
As for practical reasons (I assume you mean rewards), the players will be getting rewards while they explore towards their objectives (and anywhere else, but likely more towards the objectives as the players couldn't be the only ones with those objectives). There are many more people in the research base than them, they aren't a small crew. As a military research base, many of them likely would have grabbed nice stuff before they got cut down. Some might have even survived somewhere. And some objectives might even be right inside somewhere with goodies. And, heh, I might make it near impossible to escape without getting at least some gear and difficult even then.
Staying alive is already a primary objective. Why else would there be a rescue incoming? As for killing everything, that would be the antithesis of surviving (enemies may be hard, and often may be numerous) and I would end the game anyway once the rescue ship arrives and leaves (and yes, it will leave without the players if nobody is there for a while), meaning they wouldn't get the chance most likely.
Also, I am fully okay with people forming their own objectives out of the blue, like getting weapons from an armory or destroying a random room because they can (though destroying the only room out... heh, may be a bad idea!).
As for formal ways of cancelling, well, possibly, but I don't want to railroad everyone's character more than their character would be if it were in... "reality". Their characters wouldn't be forced by an outside force to finish an objective if it were reality. Honestly, railroading at all besides rescue is not in my idea of the game at all.
Actions will be spoilered by groups of people, which will hopefully be not read by players in other groups on their honor, unless it's secret actions/secret conversations with others, which will instead by made by PMs. If a player makes a secret action in a group of players, other players will have to roll observation rolls to see the secret actions. And if they spot them, they'll either get a spoiler for themselves, which will either once again be an honor spoiler system, or perhaps PMs if people abuse it or if people don't trust each-other's honor.
I'd probably just go to PMs directly instead of faffing about with potentially critical information everyone's on the honor system not to see. Either way, beware the issue with hidden actions- it's not nearly as fun to read.
For honor between groups, it honestly isn't likely to be critical. If it is, people should secretly preform them anyway. I'm... not going to PM the entire game, that would be too far for me. If the players mess up it's their fault. If people use it to cheat, then I'll punish those who act on it and likely be very upset afterwards (which does mean something when my games are dictated almost entirely by me!).
As for spotting, I don't really care either way. I'll likely call for a vote or something if the game should start.
I know that hidden actions aren't fun to read, yeah. Hopefully the players will expose things to the readers of their own will or give permission or something. Maybe after the secret actions are moot I'll expose what they did then. Eh, not up to me, it's up to the players. I trust humanity with some grains of salt.
I might run it, depending on how slow my other RTDs continue to be on creating them. Any ideas that would fit with this, any obvious problems, any complaints from Fniff about me distracting people, etc?
Be very, very careful with setting up the goals and mechanics. It's not enough to assume claiming your wincon and working together won't work, you've got to make sure it's just mechanically infeasible.
At times, I'll definitely be sure to make things incompatible when it matters. Or at least incompatible with those who don't think too much before they act. Otherwise, I'd be perfectly fine with cooperation even when it's unlikely! I don't like directly competitive games where there can only be one or a few winners. I want to give everyone the chance to do their objectives their own way, selfishly or "selflessly". And give them the option to skip it if they don't want to mess with others.
If, say, I give someone an objective of "shut down X", but everyone else has an objective of "take Y from X", I would be perfectly fine if the person cooperated and waited for everyone to be done. The aliens won't always be
direct enemies of humanity. If am not fine with cooperation at any time, I would state something like "shut down X, do not let anything be taken from it". Where cooperation would fail the objective. But, of course, not the game.
Role-heavy or unorthodox Mafia games sometimes have this problem, where a Day 1 full claim is better (for the uninformed majority faction) than keeping your role secret like normal.
Well, if that happened, I would most likely severely punish everyone who did so and very likely deny them the game if everyone did that. I would also remember it for any of my future games.
Also, like I said, people may get objectives to murder or stop others. Where if it were obvious then there would very likely be conflict there.
So again, run a few tests or hypotheticals breaking your own system. If everyone suddenly decides they're going to trust each other completely and work together completely, and it works, that might very well happen when you run it for real.
Yep, I'll check that.