If you receive a transaction on your node, and it is confirmed, you will regard it as a good one. The coin being spent may not exist on the other chain, or there may be another transaction on the other chain spending it differently. So you will lose coins, and can easily be scammed.
That's not what I'd call "losing coins because you run a node".
Great. I'll tell pepole who lose their coins that you have chosen a different name for it, so it must be OK.
Again, nobody CAN lose a coin just by running or not running a particular non-mining node. That's trivially obvious.
Yep, so you will make two coins. And if you were to do that for every minor improvement, we would have more than 20 different bitcoin blockchains already, and most people, including me, would see it as a failed experiment.
Why ? After all, each time you make two coins, they are evaluated in the market. If most users find the improvement much better than the original, they will put all the market cap on the new one, and none on the old. In PoW, the miners follow the market cap (in fact, they follow the block reward in $$ but if the rewards are comparable, the miners follow the market cap). So if users decide to buy up the new coins, and dump the old ones, the market cap of the old ones goes to 0, and the market cap of the new one takes over all of it.
If the users are divided over the utility of the modification, two coins emerge with comparable market cap, mostly half of what used to be the original coin, which is a good thing too: it means that essentially a useless modification was applied, at least as evaluated by the market.
I would think the multitude of different coins a good thing ; but actually, one doesn't need to fork off bitcoin to do so, one can also start a chain from scratch. That is what about 700 or more other crypto currencies have done. The more the market is varied, the more different coins there are, the better the experiment, and the more fluid the market ; the better the outcome. Monopolistic markets are a bad thing. Variation and choice is good in my book. I think bitcoin has been having a monopoly for far too long a time, and finally, other crypto is being considered somewhat, even though in a crazy speculative game.
If there is a strong majority forking away in bitcoin, the slow difficulty adaptation in bitcoin will simply kill the small minority chain: its block rate will be ridiculously small. So any 10/90 % split in bitcoin will result in a 0/100% split almost directly by the miners, and no miners are going to stick with the slow chain ; that was the difference with ETH/ETC.
Note that you are contradicted that every hard fork needs to lead to two prongs in reality: most regularly hardforking coins like ethereum, DASH, monero, .... are in any case so much centralized on their "Core" dev team, that only in very contentious cases, two prongs emerge, as was the case with the winding-back HF on ethereum.
Of course, because segwit is a soft fork. Hard forks don't work that way. The blocks produced on the other side won't get orphaned. Both chains will continue as different coins.
Yes. That's good in my book.
Yep, and if someone want to do that with bitcoin they are welcome. Just don't call the new coin "bitcoin", because that will be confusing to everyone, and don't expect anyone to use their new coin.
The funny thing is that a UASF is exactly the same. You make two coins if the soft fork chain is minority, even though it is a soft fork. UASF is asking users to only consider a segwit-only chain, and hope that a minority of miners will make a segwit-only chain that those UASF nodes will accept (otherwise, they stop). But the old majority chain continues of course. We ALSO have a fork and two coins: don't call the new one "bitcoin" in that case, you might confuse people.
However, this kind of fork is extremely dangerous. Because the majority chain may be orphaned after a month or so if they don't transform it into a bilateral hard fork.