The Chernobyl reactor system did catastropically fail. Yes, this was foremost a result of human stupidity. That doesn't change that it failed; it had several inherent design flaws that led to worsening the casualty. Would the reactor have been safe otherwise? I can't answer that question; but perhaps the casualty wouldn't have been as severe.
Most notable of these design flaws were the boron control rods and the rod followers, the positive temperature reactivity coefficient (also called a void coefficient), the excessive steam piping and design complexity of said piping, and the design at-power refueling pressure housing.
Yes, Chernobyl was one of the safest reactor plants at the time, until the operators decided to go all rogue on the safeties; but that didn't make it a good design. In fact, one could say it was because of that false sense of security that they proceeded as they did.
The operators at Chernobyl-4 believed the reactor to be so safe, and had so much faith in it, that they did not believe a major accident could occur. Wanting to proceed with the experiment (because said engineers had noted the massive deficiency of the diesel generator backup system not providing DHR flow for upwards of about 90 seconds after complete loss of power), they disabled all of the safety interlocks.
They planned for the experiment to occur on April 25th, during the day shift, and like all nuclear things, it got massively delayed and didn't start until the early morning of the 26th, with an extremely junior crew. They set themselves up into a bad condition where they wound up with extremely low reactor power densities concentrated in the core. The reactor test was performed when Chernobyl-4 was scheduled to be transferred offline for a maintenance shutdown and the test was proposed to occur in tandem with that, maximizing power production. However, the shutdown was delayed further still due to wanting to maintain power grid production for peak-hours.
The test called for them to be at a power level less than a third of the typical operating power level (low power levels being extremely dangerous areas of operation for the RBMK reactor type), the relatively junior operator inserted rods too far and dropped reactor power levels to near shutdown, and concentrated the neutron flux in the bottom of the core.
To compensate for the xenon preclusion they experienced as a result, they overrode the safety interlocks of the computer system for rod control (Chernobyl actually had a very advanced rod control system (actually two independent systems, an automatic and a manual one) involving rods that were pulled out of the top of the core and pushed into the bottom of the core allowing for severe flux distribution changes in order to allow for refueling at power), and then yanked them out, trying to get power restored.
The junior operators did not understand the condition they were in nor why reactor power didn't restore as a result of their actions manually overriding the rod control systems. In addition, they fired up more pumps to raise the flow rate, setting up what in essence was the precondition for a massive cold water casualty - one of the worse casualties that can occur to a PWR/BWR plant.
When they attempted the test, water flow was severely reduced as the turbines sped down off their regulators, causing the formation of steam voids inside the core; that combined with their positive coefficient of reactivity led to a runaway reaction rate that eventually overwhelmed the xenon preclusion (via burnout) they were experiencing as a result of initially driving power down too low.
No one really knows why the operators decided to scram when they did, reports go both ways of operators noticing the massive power transient or simply having concluded the experiment, they attempted to shut it down. However, one of the design flaws mentioned earlier also came into play here, further compounding the casualty.
The scram that should have shutdown the core actually increased the reaction rate, leading to core meltdown (if you can call it that, the meltdown at Chernobyl-4 was extremely energetic). The core meltdown and subsequent violation of critical heat flux led to a massive steam void formation that blew off the lid of the core; since the lid was rated for a meager pressure transient as it had to be in order to allow for the at-power refueling of this particular type of reactor. The 90 degree turn of the piping and the fact that water likes to take the easiest path further led to complication and expounded the casualty.
Then the real explosion occurred, that being from the hydrogen generated in the metal/cladding/water reaction as a result of fission product release and fuel element failures combined with the massive heat and power distribution. That explosion is what sent chunks of super-heated fuel plate material into the atmosphere.
Granted, the stupid humans shouldn't have removed all of the engineered safeguards. Same thing with all of the other reactors that have failed. All of them had wonderful safety features, just human beings decided they knew better and wound up breaking the plants.
No matter which reports you read, that of SL-1, HTRE-2, HTRE-3, TMI-2, or Chernobyl-4, or the multitude of mishaps/accidents involving nuclear fuel (such as the INES level 4 incident in the Ibaraki Prefecture), they all share one thing in common: the stupidity and ignorance of humans who lack the proper respect for what they're playing with.