Yah. And let me guess--most of your colleagues don't have the technical skill or knowledge to follow or construct a proper experiment. Someone set it up like this once, published, and now it's in the literature as a method for everyone to copy without critique or understanding. It's even worse with statistical models and tests where researchers pass around R and SPSS scripts for turn-key data analysis because no one in the social sciences (and most in the life sciences (and sadly many in the physical sciences)) understands enough stats to fully design and understand their experiment. And then they scratch their heads over widespread reproducibility problems.
No... you understand this really, really well. As far as your last paragraph, I have a lot to say here, but I'm currently on my phone waiting for a student to show up for a make-up exam, and typing it all out would be mad painful. I'll come back with more whisky in me and a better keyboard...
The net was loosened to simply to produce more phd graduates..
then the university can point to them and be like, look at all these smart people we've produced..
It's like that but also different. You have to reason from the top, which consists of corporate psychopaths who are merely in it for money, status, and power, but in sicnece mostly the latter two (you really don't get rich of science mostly, exceptions apply).
Hiring incompetent PhD's who have no clue how science works, don't make it to the top. So the top is safe again. I knew I wanted to do research from day 1 as an undergrad. So I affiliated myself with people who would "transport" me into a science position. I always looked for mentors from whom I could learn the tricks of the field. That brought me very far and gave me a competitive edge. Now the people at the top in my field at my organization don't like me "coz I know my stuff". Everytime you are attacked for your "research content" it's merely a political game. Things like "Stay away from my position you threaten me" becomes things like "your sample size is too small" or "that idea is so straightforward why research it". You know same thing like suppressing citizens becomes "because of national security reasons".
Further, the top makes it increasingly different to get there. Picture this. Back in the 90's, top journals didn't had these very strong data requirements. So two things are going on here. First, of course the way experiments were designed and data was collected and analyzed wasn't stringent enough. Yes, it was too easy to exclude cases, to get away with small sample sizes, and not replicate your studies and get in top journals. That's all true. But now that is used as an authority argument for the top.
The people who are now the top are those people who got publications into the key top journals with such sloppy science, who are now telling the new generation that they suck and impose ridiculous data collection (empirical) requirements that nobody has the resources for. Win-win: those new generations don't get to the top, because the requirements are so stringent you won't get to the top because you won't get your top publications.
Finally, what this does is, the system makes psychopaths. If you want to reach the top, there is only one way: cheat. Fix your data. Remove cases. Create the perfect dataset. STILL top journals after all this fraud don't want complex statistical analyses. If your story isn't simply or straightforward you won't get in. Similarly, if you don't fit their political agenda (for instance you research theory X but theory Y is popular over there), you also won't get in.
All these "do good science with good data" is just window dressing for the top to keep themselves safe in their castle and keep everyone else out.
The days when "science was all about working on interesting things" are long gone. The culture in, at least, social science and economics (which is social science) is exactly the same as lawyers, politicians, and accountants. The currency is academic papers, positions are scarce, and it is one big competition.