This article has some good points and some bad ones too.
e.g. Academic papers violate the "bottom line" rule.
If you were making stuff up as you write it down then it would be true that starting with a conclusion might bias your arguments towards justifying that conclusion. This might describe how a (bad) journalist writes, but not how scientific papers are usually written.
The majority of the work for a scientific publication usually comes well before writing the first word. Theorists will discard many theories before they find one promising enough to do extensive work (primarily aimed at proving that theory wrong) before they even consider publication. Experimenters must work with theories that have been through the above, conduct a complex (and usually expensive) experiment, and then analyze the results extensively before they start writing. The writing process can trigger some consistency checks and digging through references may reveal new spins for or weaknesses in the theory or experiment. However, if the entire point of the paper was wildly misguided to begin with, it's unlikely writing the paper in a different order (so that the authors aren't beguiled by their conclusion) will be what reveals it.
Now, the big benefit to writing scientific papers with the conclusion presented right up front is that you don't have to read entire papers to figure out what they're talking about. Scientists generally don't read papers to be led down a grand story with a surprise payoff at the end. The first thing they want to know is what this paper has to say so they can decide if it's worth reading. Science is big. Really big. Go to arxiv.org (a preprint repository for both articles that will be published and some that won't). Pick one specific sub-field and look at how many papers were submitted yesterday. The reason few people read material from outside their bailiwicks is because they barely have enough hours in the day to keep up with their own field and do their own work. If scientific papers followed the "bottom line" rule things would get much, much worse.
Personally, my biggest complaint about journals is how expensive they are. The only reason university researchers have "free" access to journals is because the universities pay (frequently exorbitant) subscription fees to the journals. It's better than having every researcher pay $30 every time they want to look at a paper, but it's still far from cheap. The really weird thing is that it costs several thousand dollars to publish a paper in a decent scientific journal. Yes, you do ground-breaking work, write a beautiful paper about it, and then you have to pay the journal to publish it if they accept it, and color figures are almost always extra! Where does all the money go? Journals have remarkably few expenses. In terms of producing content the only people they really have to pay is their editors. The authors pay and the referees work for free. Publishing the content does require some infrastructure, like servers and code. The print version requires type-setting, production, distribution, etc... However, I find it very difficult to believe that journals need to charge what they do to either creators of the content or the consumers of it. I'd love to see journals open up their books and reveal where the money is going.
e.g. Academic papers violate the "bottom line" rule.
If you were making stuff up as you write it down then it would be true that starting with a conclusion might bias your arguments towards justifying that conclusion. This might describe how a (bad) journalist writes, but not how scientific papers are usually written.
The majority of the work for a scientific publication usually comes well before writing the first word. Theorists will discard many theories before they find one promising enough to do extensive work (primarily aimed at proving that theory wrong) before they even consider publication. Experimenters must work with theories that have been through the above, conduct a complex (and usually expensive) experiment, and then analyze the results extensively before they start writing. The writing process can trigger some consistency checks and digging through references may reveal new spins for or weaknesses in the theory or experiment. However, if the entire point of the paper was wildly misguided to begin with, it's unlikely writing the paper in a different order (so that the authors aren't beguiled by their conclusion) will be what reveals it.
Now, the big benefit to writing scientific papers with the conclusion presented right up front is that you don't have to read entire papers to figure out what they're talking about. Scientists generally don't read papers to be led down a grand story with a surprise payoff at the end. The first thing they want to know is what this paper has to say so they can decide if it's worth reading. Science is big. Really big. Go to arxiv.org (a preprint repository for both articles that will be published and some that won't). Pick one specific sub-field and look at how many papers were submitted yesterday. The reason few people read material from outside their bailiwicks is because they barely have enough hours in the day to keep up with their own field and do their own work. If scientific papers followed the "bottom line" rule things would get much, much worse.
Personally, my biggest complaint about journals is how expensive they are. The only reason university researchers have "free" access to journals is because the universities pay (frequently exorbitant) subscription fees to the journals. It's better than having every researcher pay $30 every time they want to look at a paper, but it's still far from cheap. The really weird thing is that it costs several thousand dollars to publish a paper in a decent scientific journal. Yes, you do ground-breaking work, write a beautiful paper about it, and then you have to pay the journal to publish it if they accept it, and color figures are almost always extra! Where does all the money go? Journals have remarkably few expenses. In terms of producing content the only people they really have to pay is their editors. The authors pay and the referees work for free. Publishing the content does require some infrastructure, like servers and code. The print version requires type-setting, production, distribution, etc... However, I find it very difficult to believe that journals need to charge what they do to either creators of the content or the consumers of it. I'd love to see journals open up their books and reveal where the money is going.