You're almost right. But you have to do the job to the end and do the test with real accounts. Demo accounts are not relevant. Often the price stream is not the same, and anyway, on a demo account there is no slippage.
I made this effort!
I currently have a real account with capital by the following brokers:
• FxOpen UK
• IC Markets
• TickMill UK
• Squared Financial
• LiteForex EU
And the list of brokers where I had accounts currently inactive and still much longer.
And as you say very accurately, the two tests that must be done is the same EA deployed at the same time on all accounts, and the best is a master account who send the same trades on all accounts in slaves.
This makes it possible to rule out the worst brockers for such strategies.
But once it remains the best, things are very complex to test. The results are not always stable and linear. It would be necessary to test for months to have really valid results.
Some brokers are better on some pairs at certain times, etc.
Darwinex are not always the best everywhere, but they probably have the best overall results.
And if we weight the results with the 'pure agency' model and the long-term security that this model offers, Darwinex is objectively one of the best MT4 brokers that exists.
The fact that you do not compromise on the quality of the brokerage to have access to the environment of Darwinex Asset Manager is really great.