wavelets as it's own chapter, deep learning only has GANs as a single subsection, graphical models several chapters after the ML chapter...just weird arrangement choices
Also, as others have mentioned, some of the most important skills for DS are data munging, data "presentation", and soft skills like managing expectations / relationships / etc.
I would not recommend this book if you want to get into DS with the idea that, "I'll read this and then I'll know everything I need to." It's too dense and academically-focused, and it would probably be discouraging if you try to read this all without getting your feet wet.
Most data scientists are consumers of algorithms, not producers of algorithms. The rules are a bit different if you're at a bigco, but most data scientists don't do active research. It's nice to have a solid theoretical understanding machine learning, but most data data scientists' day-to-day consists of chaining together libraries and building nice dashboards.
This - though I'd add that data science is also about statistical intuition, knowing what questions to ask of the data and how to get sensible results. It's not essential to understand that PCA involves finding the eigenvectors of the covariance matrix, but it is essential to know when PCA would be useful, and common gotchas that might make it irrelevant.
Is there anything that this book would be missing the day to day stuff?
But really understanding the failure modes of what you've made is much harder than that. Unfortunately I don't have a magic bullet or even a reference for that side of things. It's just something I've learned on the job. To paraphrase the Anna Karenina principle :
All good models are alike; each bad model is bad in its own way.
Anyone who believes this is dangerous and shouldn't be allowed anywhere near a data science project. The programming is easy, but working with data and making good inferences is very hard.
But I understand what you mean, and completely agree that poseur amateurs who treat DS as just easy "number fluff" and expect fancy ML frameworks to solve all their problems are cancerous and ruin the reputation of DS as a field.
Yet the more I have worked with regular people at work, the more I have moved from the camp of always doing the "right thing" to hoping that people would just "do something". I can't fix every problem so maybe I'll just deal with the reality as it is and try to make best out of it. I won't start quizzing people at work about DS know-how but maybe then silently guide them towards understanding what they are doing instead of driving them out of the room and keeping them located somewhere far away from the data science team.
Knowing the theory is important, but most of the time what you actually need know is to quickly knock together a script that pulls in, cleans up and merges some dirty data from 3 different sources, select the right out of the box algorithm for the situation and presents the results in a clear and pedagogical way.
So no, you'd would be completely unqualified. You would however gain a deep understanding that might help you come up with novel new techniques for solving large scale problems.