Thank you for your comment! I agree there are many ways to approach this, and conclusions depend on the specific situation. My goal with this post isn’t to discourage dictionary use in general. Instead, I’m highlighting certain important reading skills, like guessing from context, that need practice. Immediate translation doesn’t usually help build these skills.

I used older research mainly for convenience and clarity, but there’s plenty of newer research supporting inference and context-based guessing in second language learning. Unlike the whole language vs phonics debate, which deals mostly with first-language reading instruction, my focus is on adult learners of Chinese on all levels. From what I’ve read, in second-language contexts, meta-cognitive reading strategies are generally regarded as being critical for L2 reading. Here is a fairly recent meta-analysis on L2 reading strategies: Yapp, D. J., de Graaff, R., & van den Bergh, H. (2021). Improving second language reading comprehension through reading strategies: A meta-analysis of L2 reading strategy interventions. Journal of Second Language Studies, 4(1), 154-192.

Over-reliance on dictionaries is also a real and common issue I’ve observed teaching Chinese. Students often feel anxious or stuck when faced with unknown characters without immediate translation. My suggestion isn’t to stop using dictionaries entirely, but rather to occasionally practise reading without one or limit its use.

Finally, it’s certainly possibly to avoid crutches, or at least limit their use a lot, by focusing on extensive reading of easier texts. 

The reading stategies aren’t a good reader method, they are a poor reader method and are also a crutch. Of course in second language we will generally initially be poor readers because in most cases we don’t learn to read a language which we already largely know in spoken form but do everything concurrently. So in the end we need to rely on some crutches. So the discussion is from the initial point should be which crutches are more advantageous, but this values one over other without supporting this with any arguments.
Is being able to actively guess the meaning in situations where you can’t actively check it better or is quickly learning the meaning that will likely be the correct one instead of risking remembering incorrect guess better?

I would bet that this largely depends on the situation you are in.

But it’s kind of nonsensical to reach for old research here, since they can’t compare the situations. You should also address it directly and not just throw a list at the end.

Yes, I usually recommend that too, not only for this reason, but because it’s usually not necessary to learn a word unless it occurs a few times in a text. Learning everything is always a bad idea, unless the text is tailored for the specific student in question.