>no human can have such visual memory how to know every single angle?
I dont think you got the point of what I meant, but let me clarify myself since what I said was kind of lacking anyway.
What I mean is that, when you cant find a precise reference for something, you can simply create it yourself by grabbing as many useful images as you can (keeping them near you for quick consulting, I mean) and making the combinations for the whole composition inside your head while also changing some things to make the combinations make sense.
An example of this would be using three different draws to make a final draw based on A's legs, B's arms and C's facial expression. You acquire the references for each element you need separately, then combine them in your mind to have the general idea and accurately make the connections, while you have the original references at hand to for consulting the details of each element separately.
This method has helped me a lot whenever I have problems finding a reference, it's also a good way of studying poses and construction as you're essentially doing what OP did but in your head and making the corrections there and on the canvas. Granted, this method gets more difficult to pull off depending on the elements you need to combine, and the incompatibilities can be a pain to deal with too.
To answer your question, no, Im not talking about visual memory here, it's about using references and mental construction.