In docs for both methods:
https://docs.mongodb.com/manual/reference/method/db.collection.updateOne/
https://docs.mongodb.com/manual/reference/method/db.collection.updateMany/

you can see:
> Using the update() pattern of field: value for the update parameter throws an error.

This is example how mongo shell handles it:

> db.test.insertOne({_id: 'test', field1: 1, field2: 2});
{ "acknowledged" : true, "insertedId" : "test" }
> db.test.updateOne({_id: 'test'}, {});
2017-03-24T00:18:27.129+0300 E QUERY    [main] Error: the update operation document must contain at least one atomic operator :
DBCollection.prototype.updateOne@src/mongo/shell/crud_api.js:519:1
@(shell):1:1
> db.test.updateOne({_id: 'test'}, {});

Then if you try to do updateOne using node driver it will not throw error:

var testCollection = db.collection('test');
testCollection.updateOne({_id: 'test'}, {});

and object will be updated and will be looking like this
{_id: 'test'}
other fields will be removed.

This is dangerous operation and if client code made it by mistake it can lead to big problems with data. So, I think updateOne and updateMany should follow reference and throw error in this case.