Staying in a lane though is more than just gold, it is also war score and opportunity for further kills. It is pushing a lane, taking down towers, etc. You are sacrificing all of these things by dying, just to get a little more gold and XP.
Maybe you haven't noticed, but the team that wins isn't always the one that has more levels and a bit more gold. It's the one that kills the citadel first. While they often go hand in hand, risky strategies to get a fairly mediocre advantage are not worth it.
So let me condense this down into a few clearer points:
1. Kill swapping works on two bits of information: You think you can kill the opponent, you think you will die trying. You are almost alwyas right about the latter, and much less likely right about the former.
2. You will get more gold and xp than the other team, in MOST cases, but not all. If you solo kill, and two enemies get assists, it equals out.
3. You can often drive someone off, without increasing the risk to yourself, much more readily than going for the kill swap. This is the cost of the kill swap, not just the money/xp any opponents get. It is an 'opportunity cost.
4. Being alive allows you to push lanes, destroy towers, and put your team into positions to gank the reamining DG's.
5. Your effect while being alive may be greater than the effect of the opponent while being alive, and dying will hurt your team by the difference between these two abstract numbers.
The team with the gold and XP advantage will win the vast majority of the time. Especially if they establish that advantage early (which is probably when most kill exchanging happens in my experience). That makes trying to get a gold/XP advantage a very smart strategy.
Over half the time your team will have the flag anyway. You either have the flag already or it was a 2v1 (which is why they were running) and your teammate will get the flag. And if you are giving up gold/XP advantages to get time to damage a tower a bit more, well, I have to question that decision.
In conclusion, I dont see any rational way that you would want to exchange a small amount of warscore (on average) and the possibility of doing a small amount of damage to a tower for 1k+ gold and a solid chunk of XP. That exchange you are suggesting doesnt make any sense to me.
Point 2 is wrong if one of the assists is the guy who is now dead. The people getting the assists would have to be 2 nondead people, which is a rare situation. Plus it doesnt even out since I think you will come out ahead in terms of XP, while the gold is even (somebody can check me on this, not sure).
What you are missing from all your "you need to be alive to have the possibility of doing something else" is that by killing your opponent you also remove them from play for 30 seconds or so when they probably could have been back in play much more quickly by simply doing a quick crystal run and tp back.
I'm not here to change anybodies opinion (including my own) nor are any of you here to have them changed in first place. No matter what, I'll never die for a kill and I'll do my damn best to never die, period. As long as I'm not feeding kills, nor are any of my teammates - we'll win. Following someone into their base for that last hit, or not.
Not to be mean, but you should be here to change your opinion. The whole reason to read strategies is to change your mind about how to play the game. Why are you reading a strategy board if its not your intention to change your reasoning?
We have suggested detailed ways in which kill exchanging is better. All you have done is repeat the EXTREMELY simplified saying that "I'm going to try hard not to die" and not stated how your strategy will work better than intelligently pursuing exchanged kills. The "dont die no matter what" is great advice for a newbie. A newbie shouldnt be trying to kill exchange. But you have to accept that once you get better at the game, some of that simplified advice for newbies is going to get replaced by more nuanced ideas.