Doesn't this encourage inefficient coding?

[brad4d]

One of the examples in this gist is:

---

 if (node) {
   node.kind = "init";
   node.type = "Property";
}

could be rewritten as

node?.kind = "init";
node?.type = "Property";

---

This is performing the same test twice at runtime.
The original was more efficient.

[nicolo-ribaudo]

This does not encourage neither more nor less efficient coding. In the example you quotes, it's true that there is one more nullish check but, in the case of a non-nullish node, there is one less variable lookup.

In some other examples, such as if (element.node) element.node.decorators = null;, using element.node?.decorators = null avoids reading element.node twice.

[MadProbe]

Also JS engines could be smart and hoist the same conditions if it's safe to do so, so this is not very concerning of an issue.