How to handle the weights and benchmarks (re-generate or switch to substrate defaults)?

[clangenb]

This is only a testnet, so we don't really rely on accurate weights. Now we have some very old weights (from 2022!) and some substrate defaults (=()).

I suggest to either benchmark the weights properly again or entirely switch to ().

My preference is (), any objections @brenzi ?

[brenzi]

I agree that weights don't need to be accurate on the testnet. but if we remove them, there's less we're covering with our testing, which I don't like

[clangenb]

Ok, then the conclusion is that we fix our benchmarks properly XD