Well, thats a bit much to answer, so in small chops:
So another case of "aircraft flies into big cumulonimbus, gets ripped apart"?
Maybe. Possibly not.
How much do airplane radars see?
That file should be a very good resource there for you to get an idea:
http://www.airbus.com/fileadmin/med...ems/AirbusSafetyLib_-FLT_OPS-ADV_WX-SEQ07.pdf
Especially note the tiny detail, that applies to any modern weather radar despite all technological advances: Shapes tell you more than colors.
Aren't the pilots trained to avoid storms if at all possible?
Usually, its OK to assume that you can fly over the storms, if you are in cruise altitude. But that does not apply to tropical thunderstorms, which can reach 50% higher than thunderstorms in higher latitudes.
With Cbs as high as they get in the equatorial region, overflying them should have been a no-no for most aircraft. And ending up in the upper region of a storm cloud and close to the service ceiling must be nasty.
The problem is: You have not just one thunderstorm. You have many. And as the file above describes well: The black calm behind the storm can easily be another huge thunderstorm hiding in the shadow. You can't just navigate around them easily. You often have to fly right through them because you have no better path that you know with certainty.
And then, you also have to include the characteristics of large Cb clouds - a clear notch in the cloud can really mean free air. Or the most turbulent place filled with ice you have ever seen. As the article by Airbus already explains: Ice has a lower reflectivity as rain: It would appear with less red colors than actually safer regions.
But in this case, you still should remember: Many more planes flew through that region without problems. And the plane was constantly flying much slower than it usually should have been flying. Its thus pretty much possible that the accident is way more complex and technical than just "aircraft flies into big cumulonimbus, gets ripped apart".