You seem very emphatic about it. Why can space travel NEVER be as cheap or as safe as air travel? SpaceX is certainly doing a lot of work towards getting costs down, and you didn't specify which particular cost metric you're using? If it's $ per passenger-mile, I think a crew dragon going to ISS can get pretty close to early air travel costs, if you allow the costs to put up ISS to be sunk. (And overhead associated with keeping ISS livable is also much lower using commercial space launches that it was under the Space Shuttle.) And modern airports aren't exactly cheap...
There's a couple of big reasons.
1. Space is
really hostile to life. I mean
really fucking absurdly hostile. There's no oxygen (3 min) and it's kinda hard to get rid of your carbon dioxide. There's no water (3 days). If you are operating, it's hard to get rid of your waste heat. If the power isn't running, things will get a wee bit chilly. There are periodic bouts of absurd amounts of radiation. There's no easily accessible source of food or supplies once you leave Earth (3 weeks). And to boot, getting to the one known safe place involves extreme temperatures on the outside of the spacecraft that is protecting you. The times in parentheses are about how long you have to solve the problem before you die.
You're going from an environment where you have oxygen within 3 minutes, water within 5-10 minutes, and food within 4-5 hours, to one where where all of these things are (at current technologies) hours (LEO), days (cis-lunar), and months (cis-martian).
Here's an example:
One of the challenges that is an unaddressed known risk for a trip to Mars is that a coronal mass ejection will generate lethal amounts of radiation over a large swath of space. If you're downrange, you've got just a few hours to get clear, and a whole lot of distance to get clear of. The plans for mitigation I've read about are: carry a bunch of water (mass) and have everyone hide behind it.
Look at the critical things above: food, water, air. Any system associated with those is fault critical. Likewise anything that deals with navigation, thrust, or thermal conditioning is fault critical. OK, now any system that supports the things above is fault critical. As in it fails, you
die.
To mitigate (not eliminate) the risks, it means that the equipment has to be designed to be very reliable in very extreme environments, and you need to carry multiple redundant sets of equipment and spare supplies. And then you get to do a lot of praying that you don't have common mode failures.
2. At the beginning of the last century, a notable science fiction author determined the equation of state for reaction engines (any and all). This is known as the Tsiolkovsky rocket equation.
M
i/M
f=e
deltaV/Ve
M
i is how much mass your rocket has to start.
M
f is the amount of mass to deliver to your destination.
deltaV is how much you need to change your velocity. For low earth orbit, this is about 9 km/s.
Ve is the exhaust velocity, i.e. how fast you can throw mass out of the back of your rocket.
The current engine technologies give us a Ve of about 3.2 km/s.
Plugging that in gives that our initial mass needs to be about 11X what we deliver. Or, for every kg we deliver (spacecraft and cargo) we need 10 kg of fuel and oxidizer.
That's a lot of fuel and oxidizer, compared to the structure of the spacecraft. And since you have both is close proximity, there's really not much separating you from a wanton orgy of devastation.
Also, when you light off the engines and start chucking mass out the back at high speed, you get a lot of vibration, and there's really nowhere for it to go, so vibration loads are very high and thus fatigue damage is also high.
Point 1 means you need to take a lot of stuff with you to make sure that you can continue operating after one or more failures in a given system. Point 2 means you don't have a lot of mass budget with which to do it in. So, that means that everything has to be heavily engineered to shave off weight which means high cost, and also means staring hard at the safety margins and trying to trim as much fat off of them as possible, which means a lot less margin for error.
Years ago, when I first became really into space, the thought was that $1000/kg to LEO was about the limit of what we could do. SpaceX is currently at ~$2700/kg. At $1000/kg, you're looking at a minimum cost of about $100,000/seat.
Let's put it this way, right now, there are no flying rockets that can fly and safely recover after critical engine out at launch. Astra barely avoided a RUD on pad when they lost one.
Every commercial aircraft on the market today has this capability proven.