Firstly, thank you for your appreciative comments. I was very pleased to have been asked to write the article and doing so was my pleasure.
You make a very good point and you are absolutely right. It should have occurred to me that the defining characteristic differentiating the two arrangements would be the 'pot' element. Thus, I am not making a fair comparison. I think I went too far there in an attempt to make the point.
Without in any way attempting to make excuses, I will present the following.
A footnote I added to the original article (I don't think it made it to the published version), was that the strength of a schema can be assessed by the degree to which the elements that evoke it can be disarranged before the image no longer evokes that schema.
For example, our face schema is so strong that it's actually quite hard to arrange the two dots and curve in such a way within the circle that it does not evoke the 'sense' or 'feeling' of 'face' (I actually had to move the elements outside of the circle to stop it from doing so).
Thus, with arrangements of element that evoke schema, it's not an 'either/or' event (either it does or does not evoke a schema), it's more how easy or difficult it is for the viewer to fall into the illusion, i.e. get the 'feeling' of what the arrangement represents. The ease or difficulty is an indication of the strength of the schema.
When viewing the arrangements you attach, I do feel that the lower one is still somewhat less evocative of the 'stereotypical bonsai' schema (i.e. it gives less of the 'feeling' of the bonsai stereotype).
That being said, I do notice a change in the degree of difference between them. In the way you present them, they have become more equivalent. Perhaps they both evoke a bonsai schema now. The top one by virtue of its more stereotypical arrangement and the lower one by virtue of the element suggesting 'pot', so it could be that neither really evokes the schema for tree in nature any more.
As schemata differ between people and as such are very subjective, I think we'd have to put it to the vote to test it. Nevertheless, you are quite right in making the point.
Anyway, in answer to your question 'how does this relate to why we see some trees in nature and think of a bonsai'? I think it relates to why we see some trees in nature and think of a bonsai because the degree to which this happens can indicate the presence and strength of our schemata.
Participants in this forum are likely to be intimately acquainted with trees and bonsai and so will have very rich and complex schemata. Thus, their richer and more complex schema for 'tree in nature' is likely to be broad enough to encompass also the more stereotypical arrangement I presented to evoke the bonsai schema.
For example, a possible reason for my including a 'pot' element in that image is that without it, the arrangement evokes in me the 'feeling' of the willows growing by Hampstead ponds (as a part of my overall tree schema). So, for myself, I had to include the 'pot' element in order for the arrangement to evoke 'bonsai' (yes, I should have been thinking more).
However, it is a truism to say that people with less experience of trees and bonsai will have less complex schema. This may equally be true for those starting out in bonsai and beginning to learn. In this case, it is more likely that their 'tree in nature' and 'bonsai' schemata do not overlap and in fact may be very different.
This is where I think things can begin to go wrong because it results in the concepts and requirements for each also being very different, when in fact, they are closely related. That is, one is a representation of the other, rather than a stereotypical representation of itself as something separate and distinct from the other.
I certainly remember this being true in myself when I first started and as I said, I wasted a lot of time trying to get trees to conform to what I thought a 'bonsai' should be rather than what a tree should be.
The best way to get an idea of the existence of these schemata in people is not to present them with arrangements that will evoke them, but, as I indicated in the article, simply to get people to draw 'a tree' and to draw 'a bonsai'.
It's likely that the results you get will range from two completely different images that bear little relation to each other (e.g. like the arrangements I presented), to a complex image that could be either. For example, I think people who have a lot of experience with trees and bonsai, when asked to draw a tree, will draw quite a complex image and then when asked to draw a bonsai, might look at you oddly and simply add a pot to the existing image.
However, I think it's likely that people with much less experience of trees and bonsai, will tend to draw a more stereotypical image of a tree (e.g. something like my 'tree' arrangement), and more to the point, when then asked to draw a bonsai, will then draw a completely different image (e.g. something like my 'bonsai' arrangement).
Similarly, if we were to remove the element that suggests 'pot' in my two images and present them to a person with a complex tree schema, asking them to categorise the images, they might say something like 'that one's a willow, that one's a young oak'. But both are trees in nature.
On the other hand, if we were to present the same two images to a person less intimately acquainted with trees and bonsai (but with some superficial exposure to each), I would hypothesize that most such viewers would classify one as 'tree' and the other as 'bonsai'.
I hope this answers your question. I think it would be an interesting empirical study.