Helpful, not essential. If you are willing to use smaller engines or can be careful with the throttle (and throttle shutoff) on larger engines you can dock plenty of smaller ships.
I wouldn't want to try it with something with a 180 degree turn time of 30+ seconds.
I wonder if any of this is true in real life? Doesn't a spaghetti rocket just explode in real life?
Rocket restart is harder in real life than in KSP. Typically a rocket is designed to be started a limited number of times, this is where RCS fills the gap.
One of the big problems is that in space you not only have low temperatures you also have limited ability to reject heat (no real convection to speak of, only conductivity; which is bad because you can get heat seep into other parts, or radiation, or to a much lesser degree ablation). This means that it is harder to engineer parts for engines; you can have a heat range of +1000C to -200C and also subject often to uneven, and for restartable engines repeated, cooling and/or heating, which all leads to fatigue.
Secondly you have issues with fuel/oxidisers; cryogenic fuels/oxidisers need to be regularly stirred in low gravity because convection doesn't work the same as it would in the presence of gravity. This is why most RCS systems use non-cryogenic hypergolic bi-propellant fuels or mono-propellant fuels as the systems are much simpler; pressurised tanks, one or two one-way valve(s), possibly a catalyst, a combustion chamber and bell for the simplest of systems.
In addition to the other replies, there's also the problem of simply igniting the flame. For example, a Falcon 9 uses kerosene and liquid oxygen, which burn real nice when ignited, but won't do anything if you just mix them together and let them sit. Just like lighting a stove, you need something to get the process going. With a Falcon 9, this is done by injecting a small charge of two chemicals called TEA and TEB. They're hypergolic with each other, meaning they ignite on contact. This then starts the combustion of the actual propellants. The amount of TEA and TEB carried on board is limited, so you can only restart a certain number of times.
Using hypergolic propellants solves this problem. The trouble is that hypergolic propellants are inherently dangerous (they'll explode if they come in contact, that's the whole point) and are typically highly toxic. They're OK in small quantities, thus they show up in RCS systems, but making an entire orbital launcher using hypergolic propellants is a massive pain. It has been done, but the result tends to be more explodey than usual.
> but making an entire orbital launcher using hypergolic propellants is a massive pain.
I think Russia might disagree with you.
Wikipedia describes the Proton launcher as "one of the most successful heavy boosters in the history of spaceflight."
TEA/TEB (Triethylaluminium /Triethylborane) aren't hypergolic - they're stored mixed in a single container. They're pyrophoric, which means they burn on contact with oxygen or any other oxidiser. TEA is so reactive it will ignite on contact with liquid oxygen - very few things can do that.
The SR-71 used TEB to ignite its engines, as JP-8 was so hard to ignite.
One of the main reasons it's hard to restart a rocket in space is simply because there's no gravity to feed the propellant to the outlet of the tank.
While the rocket is running, the acceleration caused by the motor causes the propellant to pool at the bottom of the tank where the outlet is. As soon as you stop the motor, the acceleration stops and the liquid starts floating around inside the tank. If you tried to restart it now, you would get an unstable mix of propellant and pressurizing gas which might not allow the rocket to startup properly.
To restart reliably, you need to apply a force to the tank to force the liquid back down to the outlet. During Apollo, this was done with many small solid rocket motors called Ullage Rockets. These would fire for just a few seconds just immediately before the rocket was restarted. The force was just enough to get the propellant down to the outlet in time to be sucked into the turbo pumps before ignition.
Gemini also had ullage rockets. Pic related, my authentic Gemini program "Direct Ullage" indicator buttons (took me a day to find them). http://imgur.com/AeSObj9
I wouldn't want to try it with something with a 180 degree turn time of 30+ seconds.
I wonder if any of this is true in real life? Doesn't a spaghetti rocket just explode in real life?