Re-arrange datasets argument to make most sense

Description

Is your feature request related to a problem? A clear and concise description of what the problem is: "I'm always frustrated when ..."
Over the years, datasets arguments has increased everytime we add a new feature. We couldn't rearrange the order because it is a breaking change and we don't know what's the impact of it. With https://github.com/kedro-org/kedro-plugins/pull/358 merged, we finally have the freedom to change it.

Context

Why is this change important to you? How would you use it? How can it benefit other users?
Take this example, user is struggling to find the filepath option. Although in this case it may be better to have an example instead. Nonetheless, the current signatures isn't optimal.

See SQLQueryDataset, filepath and execution_options should rank before load_args. I suggest we split this into a 2-3 smaller PRs since it may be too much to change everything at once.

class kedro_datasets.pandas.SQLQueryDataset(sql=None, credentials=None, load_args=None, fs_args=None, filepath=None, execution_options=None, metadata=None)[source]¶

Possible Implementation

(Optional) Suggest an idea for implementing the addition or change.

Possible Alternatives

(Optional) Describe any alternative solutions or features you've considered.

Start with the SQLQueryDataset source and its documented signature, especially the current positions of filepath, execution_options, and load_args. Compare the dataset argument signatures covered by the requested 2–3 smaller PRs, then verify that the resulting order makes the commonly sought options easier to find.