Every mission has three parameters: complexity, resources and risks. If we ask someone to send a complicated mission with little risk, it will involve huge resource. If you plan to send a mission to Mars in such a way, you would first test hardware on Earth orbit, then Lunar orbit, then model all aspects of the target mission. With F-G situation was a bit different - we did not fly to other planets for 25 years. There was a huge amount of new hardware without flight history. Practically only fuel tanks from Fregat were previously flown.
Project itself was very complex, with a large number of instruments and many algorithms. Maximum complexity and minimum resources translated to maximum risks. It was decided the risk is ok to fly F-G. Apparently we stumbled right in the beginning of the flight. F-G flew well for two hours after the launch. It turned itself on, extended elements (solar panels), orientated to Sun and began to receive energy. It turned on all required hardware, and the Chinese satellite. It was all confirmed by telemetry. And then... it flew out of Russian zone of communication. Burn did not happen. Our communication stations are designed for deep space and could not track fast-moving object in low orbit long enough to complete long link establishment procedures.
In the beginning we could not target the satellite. It was tracked only by ground means, and position was inaccurate, calculated up to 6 degrees. That's why us, and European stations had to add modified antennas emitting wide beams. As a result we had a few sessions and received some telemetry. It said that radio module works, link with onboard computer is up. Photos of on-orbit F-G showed that it was not tumbling, meaning Sun orientation module worked correctly. We do not know when and why burn sequence was aborted. There could be many hypotheses, but fact is only one: F-G is Sun-orientated, onboard computer fulfills its function. Logic says that when sequence is aborted, F-G will await commands from ground. Apparently it is still in that mode and we will continue attempts to make it alive. As for the failure, there are many possibilities. For example it could be a programming error that could not be detected during modeling on Earth. Difference between model and real situation could be large enough to "stupify" computer. It could have also been a hardware problem. Before we lost contact with F-G we enabled power to several modules, and theoretically damage during launch (?) might have caused problems with power supply. But those are all working versions, official reason should be established by appropriate commission.
As far as we know the rocket worked nominally. However we think that launcher should have been chosen differently, not Zenit, but Proton, which could take F-G directly to required orbit. Then F-G could be turned on and verified module by module, new comms line would be tested, and trajectory corrections made. Decision about launcher was taken in 90s, and project was based on that. F-G was a sort of jumping forward (cavalry charge?) over 25 years gap. It was understood that risk was high, but imagine if it were a success! However it is necessary to work step by step, systematically. That's why we should go back to Lunar exploartion.
I would not get fixated on F-G. We have many projects. This year we launched Electro-L, a new generation sat. Our Fregat boosters have done their job on 8 launches from 3 locations, with one more launch from each location by end of this year.
Update:"Moskit" provides this translation of an article in an online journal. Author not known.
670 seconds after launch F-G separated from the rocket. It worked correctly, but afterwards F-G survived only two hours. When station "saw" Sun it should move "head" towards it. This means two onboard Sun detectors worked correctly. One hour later sunny zone ended, and those two detectors stopped working. Shadow lasted for 30 minutes, and unexpectedly F-G came out of it facing away. We did not expect it, F-G should have kept Sun orientation while in the shadow. Detectors saw Sun again, and started rotating F-G once again, but soon link was lost.
It seems that star detectors, that so many blamed failure on, never got a chance to operate. They should have been turned on 4 minutes after link was lost, but F-G stop giving sings of life earlier. Star detectors found on F-G are reliable, they work on 20 space objects for 10-12 years.
F-G was not ready for the flight. Many saw that, and many told to the management that flight control system made by Lavotchkin was not ready. Specialists simply did not have enough time to complete complex software running the mission.
People blame youngsters at Lavotchkin, but this is unfair - they did not have a chance to learn. 15 years ago when Mars-96 was launched, it was assembled in -5C (with -20C outside) in Baikonur, in a room with broken windows. These were not conditions... there was not even toilets. Spaceship failed right after launch due to booster malfunction - completely logical final.
Today we see a birth of Russian space program practically from ashes, with all the growing pains. You should have seen glowing eyes of the young people working on F-G... they only recently began to be paid decently.
If there will be repressions, all will be lost again.