Glad to hear they've agreed to the KeyMath with error analysis and better clinical observations. I also habitually record the time for each section and make notes on student strategies and test behavior, so I would also expect the school psych to do so. I think TOL c/s calculator would be quite informative. My one caution is that if you use both forms of the test in the process, it will limit re-administration by an outside evaluator for the next six months (which is the re-test interval for achievement tests), whereas if the same form is used for calculator TOL, then the other form will still be available. How important this is depends on how likely you think it will be that you will need another KeyMath within six months.
As to the higher-level math class: I think a hole in her schedule is probably not that important! However, I also don't think it would be catastrophic for her to be slightly misplaced at the beginning of the year, unless you think that this will make it more difficult to persuade the school to move her, should the test findings support a different instructional level. I would agree that using the time for assessment would make sense, though there would still likely be a bit of a gap, as assessment includes not only face-to-face testing, but scoring and interpretation, and follow-up reviewing with the math teacher.
Another way of thinking about it might be to place her in the best guess to start the year, and pull her from it immediately to do the assessment, with an already-scheduled follow-up meeting between parent, assessor, and math teacher (and whomever else is needed, depending on whether this is a formal IEP meeting) to discuss appropriate instructional level. (Say, early in the second week of school.) This would put an endpoint on the time spent possibly misplaced, give the school an item to put in the schedule (no holes!), and define when the assessment would have to be completed.
