According to During, DePalma was supposed to be an author on the paper. According DePalma, they had brief discussions to collaborate, but ultimately decided not to.
It doesn't matter because if the article is correct, it's on him to produce the raw data, there is no need to speculate. As it is, the story is fishy in every sense of the word and definitely warrants an investigation.
Apparently he doesn't have the data anymore. According to the article, independent third parties did see the data and didn't think it was suspicious. The implication of what you are saying is that if you lose data for any reason then you're automatically guilty of data forgery. Obviously that's not a great precedent to set.
It's a recently published paper, barely a few months old. Conveniently when questioned he claimed the analysis had been done years ago by someone who's now dead and the raw data is lost. I don't know about yourself but I have copies of all the raw data for everything I ever published, simply because that's the sort of stuff a researcher would naturally keep. Of course, the story could theoretically have happened as he says but it's quite the coincidence that this specific data set is now found to be irrecoverably lost after he's accused of fraud.
Simply using Occam's razor in the absence of better evidence. Which is on those making the claim to produce, you can't just publish papers and then go "I've done the isotopic analyses, trust me bro". Researchers could make up anything then, there needs to be accountability. At the very least his team needs to retract the paper.
You've conveniently omitted the independent third-parties that had seen the data. It's a messy world. Unexpected things happen. How do these third-parties factor in to Occam's razor? You need to add more parameters to your explanation as to why the third-parties would verify the existence of non-existent data.
Here's another possibility: During goes to DePalma to collaborate and asks if DePalma still has the data. DePalma says he doesn't. During sees this an an opportunity to claim credit for the work.
It's impossible to tell which scenario is more likely. Are you really willing to ruin someone's career over purely circumstantial evidence provided by a biased witness?
You mean the sort of third parties that let a paper slip through the cracks which peers contacted by Science.org claim should have never been published because of the glaring errors it contains?
What evidence do you have that anyone has actually looked into those data? In what context? You're simply repeating his claim but there isn't evidence the data was ever scrutinized. In research you're shown other people's data all the time - that doesn't mean that people actually look into it.
According to During, DePalma was supposed to be an author on the paper. According DePalma, they had brief discussions to collaborate, but ultimately decided not to.