There are lots of tradeoffs at work here:
- cryogenic stages are more difficult to design than storable liquid stages, so space agencies tend to start building storable liquid stages, then move on to cryogenics when they have some experience.
- cryogenic stages offer higher performance (Isp) than storable liquid or solid stages.
- solid stages can be simple, cheap and have very high thrust. They also vibrate more than liquid engines, and cannot be throttled or (easily) shut down early. This makes them less useful as the final stage of a launcher, when you need to achieve an exact orbit.
- liquid hydrogen (as used in the Ariane 5 core stage) is not dense, so it requires a large tank, which reduces its performance advantage
- a common compromise is to use a first stage with low Isp but high thrust (i.e. solid or storable propellant), and a second stage with high Isp. Ariane 5 does this, in a way: its solid boosters are the cheap first stage with high thrust.
- the infographic lists the early Ariane 5G. Most Ariane 5 launches now use the ECA variant, which has a cryogenic upper stage with better performance than the 5G
- Ariane in particular suffers from some decisions that were made early on to save money: the 5G upper stage used storable propellants because it was cheaper to develop than a cryogenic stage.
Ariane 5 allows us to make a pretty good comparison between storable and cryogenic propellants. The 2 versions of Ariane 5 have rather different performance:
- GS payload to GTO orbit: 6,100 kg (13,400 lb),
- ECA payload to the same orbit: 10,500 kg (23,100 lb).
This is despite the ECA upper stage being much heavier than the GS upper stage. I'll see if I can find development cost of these upper stages. One drawback of the ECA upper stage is that it cannot be restarted (another cost-saving measure), while the GS upper stage is restartable.
In general, all rocket designs are a tradeoff between:
- design cost,
- production cost,
Some launchers are designed for high performance (using expensive or finicky construction methods to save weight), others are designed for low cost (cheap to launch, despite being heavier than a high-performance design would have).