All of the evidence you are citing is still in agreement with a 10% drop rate. The key fact to remember is that you should NOT be expecting consitency. RNG is random. Random is inconsistent. Assuming the drop rate is 10%, then if you do 100 battles, there is a .5% chance that only 3 shards will drop (Anyone who wants to know how I calculated this, look up binomial distributions). Now consider how many players are playing this game, and how many battles are done over since the system has been in place. For every 1000 players who do 100 battles, on average, 5 of them should have the bad luck that you had. I am sorry that you had bad luck, and it is very bad luck. But it is not unreasonably bad luck.
Same can be said about the 40 no drop streaks. If there is a 1.5% chance of this happening everytime 40 battles are done, then it is going to happen quit a bit. But the good news is that so will having multiple drops in a row. The main point is that it does actually reduce the number of cases of bad luck, but it cannot remove them entirely. The only way to remove the bad luck cases entirely is to remove the randomness entirely. But many players have complained that the reduction of the randomness makes the game boring, so the devs aren't going to do that.