I thought that from reading the first part of the first meta sample too, but in that same paragraph is mention of a second study that apparently did find relevant issues at low doses in vitro of human cells at environmentally relevant concentration levels.

In fact the purpose of meta analysis is to compare and contrast the conflicting research and results on a topic. It's very useful when forming a scientific view.

I'm not against meta-analysis, but if those analyses rely on studies that have flawed methodologies, it is just an exercise in statistical hacking. With enough massaging, you will find something eventually.

I don't have time to check in detail; can you link the study finding issues at relevant doses?

Anyway, my thinking is that if there was such a big problem, we would have found it already. It affects the food supply of so many; it seems unlikely that there are significant issues that wouldn't show up in the population at large.

The real concern is environmental impact and, particularly, effects on insects. But since they are going to use something else that may or may not be worse, it's probably better to not ban the stuff until it can be proved that the damage is worse than the benefits…

You have not assessed the facts critically. The argument in favor of glyphosate's safety is that, as the herbicidal action is the result of disrupting an amino acid synthesis pathway that in animals does not exist, it is therefore harmless to animals. This argument is already fallacious: all it does is establish the mechanism by which it is harmful to plants. These studies evidence that glyphosate is harmful to animals and investigate the mechanisms underlying the harm. The fact that these experimental conditions are not the same conditions under which glyphosate is consumed in the food chain does not make it bad science, because science is concerned with knowledge that generalizes (e.g. biological mechanisms and pathways) and these mechanisms cannot be gleaned by reproducing the conditions already in place.

The comparison with vitamins is not relevant, and to bring it up suggests you are not thinking clearly.

To me it is you who is clearly confused. The vitamin parallel is very relevant; at the concentration used in the studies, vitamins would be toxic as well. The poison is the dose. Using dosages far above what could realistically be ingested makes the studies irrelevant. By the same logic I could prove that salt actually kills you.

On the pathway argument, you are just rambling; I'm clearly not talking about that. Whether there is a pathway is largely irrelevant if you cannot prove that it is toxic at expected ingestion levels.

You are just fearmongering and grasping at straws. Same bullshit as the anti-vax that would have you believe vaccines are toxic because they use aluminum (yes, in amounts completely benign).