Bug: Re-benchmark every pallet in runtime

[jak-pan]

As noted by @apopiak here: #188 (comment) we should re-weight every pallet in our runtime (with our settings) and use that instead of default.

[apopiak]

Just saw that Kusama also uses default weights in some places. So I guess you don't have to literally re-weight everything (that was just my gut reaction), but check all the pallets on whether they need to be weighted.

Re-weighting all of them is just the safe default. If you are confident that the benchmark will give the same result (because you use the same config on the same benchmarking hardware) you can skip it of course.

[jak-pan]

Just saw that Kusama also uses default weights in some places. So I guess you don't have to literally re-weight everything (that was just my gut reaction), but check all the pallets on whether they need to be weighted.

Re-weighting all of them is just the safe default. If you are confident that the benchmark will give the same result (because you use the same config on the same benchmarking hardware) you can skip it of course.

It's not that much work / time to do all of them on top of existing benchmarks so we might just add everything to be safe.

[jak-pan]

Also I've noticed that every transaction currently costs the same amount to execute. We should have a look at why this happens since exchange:buy should be more expensive than balance transfers. We should also check the multiplier of weigh to be in sync with current DOLLARS and later dynamic.

[jak-pan]

We should have a look here https://github.com/paritytech/polkadot/blob/5b755428f73345c0015d499482c48c37ce7b9e21/runtime/polkadot/src/constants.rs#L57