It's important to keep in mind two very important concepts:
- Launch and orbit insertion are typically two distinct phases of flight.
- Ground controllers have very limited options during both phases.
Once the launch starts, about the only thing the mission control team can do is make the vehicle blow up. Most of those fantastic explosions you see of launch vehicles blowing up are because the mission control team made that happen. It's better to have a blown-up ex-rocket than a fully function but out of control rocket that hits some city. That didn't happen in this case. The launch appears to have been nominal. The problem instead apparently occurred during orbit insertion.
ESA issued a very short press release on August 22 that merely stated the vehicles were in the wrong orbit. They issued a much longer press release the following day that stated what that meant.
From that latter press release, the satellites were supposed to have been placed in a circular orbit with a semi-major axis of 29,900 kilometers and an inclination of 55 degrees. They are instead in an elliptical orbit with an eccentricity of 0.23, a semi-major axis of 26,200 km, and an inclination of 49.8 degrees. There's not much that can be done to fix the problem. What little fuel was left in the upper stage has been purged.
My guess is that everything went more or less okay during the launch and during the first transfer burn. The problem arose during the final orbit insertion. Those numbers suggest a significant underburn at orbit insertion. Minor underburns (this was not a minor underburn) can be caused by a number of things such as navigation errors and erroneous sensors. Significant underburns such as this are more often a result of bigger problems such as engine failure or running out of fuel.
So why didn't they see this on the ground? They did see this on the ground, but it took the mission control team a couple of hours to realize that the vehicles were in the wrong orbit. GPS doesn't work at that altitude. The only way to determine what orbit a high altitude vehicle is in is to make multiple range/range-rate measurements from the ground, spread out over time. Orbit determination isn't something that can be done by one single observation.
Update:
I didn't keep track of this answer. Today is October 27th, but the root cause was discovered over two weeks ago. The root cause was loss of attitude control due to frozen fuel lines on the Fregat upper stage. Hydrazine fuel lines that fed attitude control thrusters were routed too close to liquid helium lines that supplied helium used to pressurize the fuel tanks. Per the article A Simple Plumbing Problem Sent Galileo Satellites Into Wrong Orbits at IEEE Spectrum,
The fuel — hydrazine — didn’t reach the thrusters because it froze in its feed line, which was too close to another feed line carrying liquid helium. Both lines were mounted on a same support structure that, unfortunately, functioned as a thermal bridge cooling the hydrazine. The fault has also been discovered in other Fregat stages that are now under construction in Russia, and was apparently caused by ambiguities in the design documents of Fregat.