I agree with the considerations on scope and balance from Bryan's answer. With a CS background from Germany, I consider all of the activities you listed quite normal, as long as you are not downright swamped with the sheer amount of work. I would, however, like to point out another important aspect:
It is important to learn how find synergies in the additional tasks and your research.
This can mean executing some of the additional tasks in such a way that their results are in some way useful for your research. It can just as well mean guiding your research into a direction that aligns with the content of your additional tasks, even if it is not exactly the direction you would have picked otherwise.
Let's look at your examples (without taking into account whether the concrete amount or frequency of such tasks is adequate):
- writing parts of grant proposals
Beside this being an "inter-generational contract" (PhD candidates are employed from the funding gathered by a grant obtained by one of their predecessors, and during their employment, the candidates write some grant proposals to secure funding for their successors), delving into the topic of grant proposals may give you a better overview of what questions and topics appear to be considered interesting by the community and by external entities.
Furthermore, it gives you a chance of spawning follow-up projects to what you are working on, which will, in the future, produce citations of your work. In the best case, it might even create one or two new positions in your lab with colleagues working on directly related topics, thus naturally paving the way for collaboration opportunities.
- giving popular science talks about our work on any given chance
- presenting our work during lab tours for all kinds of guests
- writing news articles and marketing material
All of these are chances to practice how to present your work to members of an audience who are not experts in your field, maybe not even in your general subject area.
Depending on your research interests, this can be immediately helpful for paper-writing: While research papers typically assume some familiarity with the field, there often is a desire to keep at least the introduction and problem description sections comprehensible to a rather wide audience. After all, even readers who may be only marginally familiar with the particular research questions should be able to figure out what the paper is about by and large, so they can decide whether reading up on the topic in order to understand the paper in-depth is worthwhile.
At the same time, chances are your research interfaces with people outside of your field in one way or another. Maybe you will have to embed a theoretical framework in a context of application from another field, maybe your research results are meant to benefit non-experts (who will thus be a part of user studies or similar events). Either way, you will have to explain to people who are outside of the target audience of your actual papers what your research is about.
- supervising students and assistants
This might be the proverbial elephant in the room when it comes to synergies between additional tasks and your research - at least if this supervision happens in the kind of arrangement I am thinking of: Projects of one or more students whose content is more or less individually agreed upon between the students and (officially) the examiner/(inofficially) the supervisor.
If that is the case, try to actively participate in the definition of the tasks. Bachelor and Master theses, as well as other student-based projects, can often be designed in such a way that they end up forming a sub-project for your own research (think a prototype with a small evaluation of a concept you describe, an alternative approach that you do not have time to explore yourself, an extension that forms the "missing link" between your concept and a concrete problem, ...). Motivated students may be willing to try and publish a paper about their work together with you after completing their project, not so motivated students will at least submit their work, which you can then mention in your thesis as a proof-of-concept developed by students.
If done well and with some good luck, this can be a self-reinforcing process, as satisfied students will recommend your supervision to (hopefully) equally capable friends of theirs who can carry on where the previous ones finished.
