I have to agree with Stuart, and disagree with whatever book(s) you have been reading. The objective of structured programming is to create code in chunks that are easy to understand (and therefore less prone to errors) and reusable. If I have a short loop, it is much easier to see what a break or continue statement will do than it is to see what some contrived conditional statement will do. The fact that the loop has more than one exit point does not inherently make it difficult to understand what it does. It could be argued that such a statement in the middle of a very large loop is a non-obvious trap, but it could also be argued that in well structured code such a very large loop should not exist in the first place,
As for the paradigm of Mathcad:
Mathcad is a whiteboard interface that is, mostly, WYSIWYG. Placement of statements is free form, and they are executed from top left to bottom right. The statements can be function definitions, variable assignments, special blocks of code such as solve blocks, etc. A special type of variable is provided, a range variable, that acts as an iterator, or loop variable. Some people, especially those coming from Excel, write worksheets where every step of the calculation creates a new vector or matrix of data, and range variables are used extensively. While that is possible, it is what I (and many others) would consider bad programming. The right way to do it is to create a series of dependent functions that carry the variables of interest to the end of the calculations, and then calculate the final result. Of course, some intermediate results may also be of interest, but the principal is the same. Range variables should almost never be used for anything other than indexing into vectors or matrices, or to define the range of the independent variable on a graph. In addition to the above, a worksheet can have blocks of text (basically, extensive comments), graphs, controls (in Mathcad 15, not in Prime) etc.
Note that in the above, I said nothing about "programs". You can do a lot without programs, and in fact there was no such feature until (I think) Mathcad 6. Programs exist because it's very difficult to do certain things without them, and even if you can it may be very hard to understand how it works. Programs can be very large, but then they are hard to understand, so a better way to structure them is again as dependent functions, each of which may be just math expressions, or small programs. The final program then calls those functions, so it's more compact and easier to understand. So programs should rarely be large, which is why the use of continue, break, and return statements is acceptable.
In a Mathcad document, most of the structure is at the worksheet level, not within any (hopefully) small program that is a part of that larger structure.
