I like the linear, parsed-down quality of the images; it makes the triptych feel cohesive. However, I think these could benefit from adding a bit of color for a focal point. You could add color to Winnie the Pooh or other main characters to draw you eye to that scene and differentiate it visually from the text.
You are giving a lot of white space for the text in images 1 and 2, which I think makes it easy to read at both sizes. I especially appreciate how the text transforms from 2 longer lines to a smaller chunk of text in image 2 — I think that was a really effective way to adjust the text for mobile. However, in Image 3, the text stays about the same, except for a few adjustments to line length. It feels a little cramped and tight in the image space on the mobile size. This is probably in part due to the fact that it is a long quote in a small space; you could try to cut the quote itself down. You could also try making the point size a little smaller so there’s more white space around the quote when it transitions to mobile.
The leading in Image 1 for desktop is a little large. I also think that you could make the point size a bit larger, as you have so much space to work with.
The contrast from light background to dark text is very effective and readable — I don’t have any issues with legibility of text. I would experiment with different point sizes and finding out how much bigger or smaller you can push it without it becoming unreadable. I would also experiment with engagement of text with the image. For Image 2 for mobile, it would be interesting to see if you could break up the two lines of dialogue to sandwich the image in the middle.
I think your typeface choice is appropriate, but maybe gets a bit too ornate at small point sizes. Overall, the image-to-type and image-to-image visual relationships and vocabulary are strong. Next steps you could take would be to engage the type more with the imagery and differentiate how the text functions from desktop to mobile.