Dear Mohammed Khalil
You wrote a perfect study, how to improve science in the future.
I will add about wrong filters for amateur papers (people not from universities). I wrote about this in my essay. One example is because arXiv almost totally rejects papers from amateurs. Their principle with endorsers leaves almost zero number of amateur papers. I agree with stronger filter for amateurs, but not with almost zero probability. They even do not allow to have references from viXra, as Phil Gibbs writes in this contest. Besides, computer analysis allows today transparency over the papers of different quality.
"Admittedly, amateurs are less intelligent in average than professionals, because professionals go through a larger number of selections. Amateurs have less time and no money for research, except their own money. But probability for correctness of their theories is not zero. On the other side, amateur's theories come across too little arguments for their rejections than professional's ones. Thus their (un)correctness is much less checked. We should know that in influential statistical parameters are mainly distributed by Gaussian curves, or at least, probability never falls toward zero. Finally, because this is only probability, it is never sure that someone will not or will and TOE or anything other. Now, there are too many over-reactions against alternative theories, too rigorous limitations who can publish, what s/he can publish, how s/he should to write, etc. Limitations are necessary,but they should be more precisely set, according to mathematical and statistical laws. For instance,natural selection respects that distributions of aptitudes are more Gaussian ones than rectangular ones, but selection of physical theories does not respect this. Thus, if we have enough large number of papers from amateurs, someone will give scientifically useful ideas. Thus, I predict that after discovery of a TOE it will appear that also some theories of amateurs are correct. Thus, maybe alternative approaches need to be respected with less prejudices."
I wish to add still your principle for publication of papers with zero results. Besides, the extreme, that some measurements are written 10 years, because of checking of results, is also not optimal. For instance one last measurement of gravitational constant take 10 years.
Besides, a possibility appeared that G measurement is changing cyclically through time. If number of such measurements would be larger, it can be statistically checked whether if this cyclical change is really a case.
One rule for the advancement of Science is also to write physical theories more clearly. For example, here is an example for a clearer presentation special theory of relativity.
My essay
Best regards
Janko Kokosar