Interesting question. First of all, I will spare a more elaborate discussion of the general problem I see with identification questions for a later and more focussed time so will concentrate on your specific problem. However let there be said, that I regard poor identification questions a major problem of this site and a possible way toward its demise. Thus enforcing a certain level of quality on those is a primary mission to tackle this problem and separate the wheat from the chaff. We can't make the error of letting slip poor and undetailed questions simply because we want to "wait and see". Which I think is the main point you're adressing here. Closing, or rather putting on hold, is not a defnite decision and a question can always be reopened if it has been improved, which does indeed happen.
First some general points about closing poor identification questions. The primary point I see in the "not enough detail" close reason is not to rule out questions that won't get an answer, afterall everyone can smash any kind of answer onto such a question and they rarely stay unanswered at all. When deciding about this close reason I not only look if the question has enough details to garner some kind of answer, but rather if it seems as if the asker has given any kind of detail he remembers or not and if it is assessible to anyone else if an answer is correct or not. Thus, I'm not looking for the mere presence of information, but also the presence of effort. This might be a more subjective and quality-oriented approach to the matter, but it is the only approach by which I think we can enforce a certain level of quality in this controversial question category before it becomes the doom of this site.
Now let's look at your examles:
Example 1: Man carries amber coloured stone and straps it to a wood staff
This question is actually a positive example as it has been greatly improved due to the asker adding more details on request. If you look at the revision history, I would doubt that there could be done anything else reasonable with the original version of this question other than close it. But in its current form I for myself would not see the necessity to close it. It may very well be that the two existing close votes on it were from that first version of the question and the people might have changed their mind now, in which case it would be a responsible action to rectract their close-vote now (of course only if they changed their minds).
Example 2: Looking for a Pre-70's B&W movie that has caves, lava, and creatures
That just isn't much to go on and I don't feel there could be any reasonable conclusion reached about its answer, no matter if some answers are given to the question or not. And even more than that, it is in its current form just not the kind of sufficient quality identification question that I would like to see here and that I think could make a proper signboard for this site. Rather than that it is the kind of random three-liner that attracts more questions of this kind and contributes to the further dillution of the brand that this site stands for. And this has to be tackled as soon as possible.
As to your general point of just "waiting and seeing" how those questions develop. As said this is not what close-voting is for. If content is to be regarded inappropriate, then putting it on hold is the right way to handle it before it causes harm to the image of the site. I always tend to keep the future development and quality of the site in mind in such cases and those kinds of poor identification questions are not what I think would contribute to a positive development in this regard. We don't want to become another lmgtfy.com, but I don't want it to become another irememberthismovie.com or namethatmovie.org either. Add to this that, in contrats to other kinds of questions, those questions are not salvagable content-wise without the help of the original asker. If it becomes an appropriate question later on, then the close-vote can always be retracted or reversed by reopening. But just letting it stay open and close it in a month is not the proper kind of action. We are not to ignore bad stuff and just let it go, but to handle it when it is still possible before the situation gets out of hand.
(However, I admit that having the broader picture and future development of the site in mind I might have a stricter quality and on-topic threshold for identification questions than for other question categories, which might come across as a bit of an "elitist" attitude. But I really deem those kinds of questions the biggest danger for the well-being of this site as an interesting place and an attraction for people interested in engaging questions. If we don't handle their impact properly, their viral nature will only lead to an exodus of avid users and interesting non-ID questions. Well, now I still taked about the broader problem, but I guess you have to whenever discussing ID-closure.)