By way of example: Literature. It's easy to read a good book; it's much harder to write one without lots & lots of practice.
Especially the "less choice" one.
Nothing prevents the staff from being retained from when legislator to the next.
Kind of like the bureaucratic equivalent of external consultants coming into a private company.
That's a little bit true but mostly misleading. As with any office with professional staff, most of the work of a legislative office is done by the staff, not the head of the office, true. OTOH, that doesn't mean the the skill, knowledge, and understanding of the head of the office is irrelevant, as the head directs the staff strategically, reviews, inquires, refines, and redirects work, etc.
> Nothing prevents the staff from being retained from when legislator to the next.
The new legislators not having the skill to select appropriate staff might, but a bigger problem isn't that the staff aren't retained but that they usually are; the less skilled the legislators are, the more the professional staff (who probably are shuttling back and forth between working for legislators and working for interest groups) do not only the detail work, but also control the agenda, and the less which faces the electorate sends to fill the nominally most important seats actually matter.
Also note that this sort of negligence is happening today despite a lack of fedral congressional term limits: https://www.brookings.edu/blog/fixgov/2017/05/24/vital-stats... . I can't find a source but I believe there is a similar issue with the per-congressperson researcher appropriations.
So while perhaps less turnover helps make for stabler staffing jobs, it's not a cure-all. Ultimately the balance between Expertise and Democracy is a difficult one that we need to do a better job at (usually having not enough of both!) with or without term limits.