In the next months I will defend my thesis and I wonder if it is really worth it to write a traditional thesis. Although I have time to write it, I would consider a “waste” of time that I could use continuing my research (or improving some of my results).
On the other side, all my colleagues have written a traditional thesis, that is why I am not sure if a sandwich thesis would harm, somewhat, my future career.
Regarding your career, don't worry at all. In computer science, it is mainly your publications, conferences, and your code in public repositories what's gonna be valued in real life. No one is ever gonna read your thesis, apart from (if you are lucky) the members of the committee evaluating it.
That being said...
In the next months I will defend my thesis and I wonder if it is really worth it to write a traditional thesis.
The question here is: do you have a thesis in the research sense of the word (a statement about the world that you believe to be true and you have tried to prove during your PhD years)? From your description, it doesn't look like that's the case.
This would be my personal advice:
Case A: You already have a few good publications that answer some questions of a common broader topic. In my experience, it is worth to compile them, cleaning them out, and adding an introduction and some conclusions to help the reader prove you know what you are talking about, and that you are able to manage the big picture. It helps you clarify your thoughts and your research. That proves that you have a thesis, you are responsible for it, and you are able to defend your own research. You might actually learn a lot doing this.
Case B: You have been publishing here and there, about different topics, some of which don't actually have much to do with each other, and you don't really have a "thesis" per se. Then, honestly, don't bother. Just present your collection of publications and get your PhD. You cannot build and improvise a decent thesis in a few months if you don't have one already.
Case C: You don't have good publications and you don't have a thesis. Get your PhD and run away as fast as you can. Try not to recommend that University to anyone in the future.
I'm a CS PhD who went into industry after my degree. When discussing my thesis back then (which, by the way, was what you call a "traditional" thesis) it was always in general terms: the ideas in it, not the pages. Similarly, any time I have fielded an application from a PhD, they would put the title of their thesis on their CV, but I very rarely would actually go and try and find the thesis. Instead I would look at their published articles in journals and conferences.
So while I can't give a definitive answer -- and I can't say that universities wouldn't scrutinize the thesis more closely and perhaps discount a "sandwich" thesis -- I'm strongly of the opinion that if your university permits it and your advisor supports it, that's what matters. The likelihood of it "harming your career" seems quite low, because in all honesty, who will notice?
It all depends on the quality of the papers and the venues where they were published. A thesis that sandwiches multiple papers from top venues is potentially beneficial, since the included material stood a particular rigorous evaluation process.
tl;dr: Writing a sandwich thesis can further narrow your focus and prevent you from developing a broader understanding of your field.
I have mixed views on sandwich theses that essentially staple 3-5 papers together between a short introduction and conclusion chapters. The benefit of the sandwich thesis is that articles are generally what is important to hiring committees and funding committees. As the sandwich does not include any sides it is also better focused and can be quicker to write letting you finish sooner (which has advantages and disadvantages).
The benefit of a traditional thesis where you have a literature review chapter, general methods, a couple of result chapters and a conclusion chapter is that it lets you explore side topics in addition to your main focus. They can be quicker to write since you only need a single literature review, method, and conclusion sections, especially if you put the writing off until the end. A traditional thesis typically has more comprehensive literature review and conclusion sections. Following side projects and writing a really thorough introduction and summary can provide you with valuable insight into the wider field and help you find your next topic. This is not to say that a sandwich thesis cannot include side projects and in depth literature and conclusion chapters.
