Eliminate multi-acquire special cases in acquisition pipeline

[cmeyer]

The special cases are mainly that the data items are updated in the ScanDataChannel and CameraDataChannel classes rather than the standard acquisition pipeline channels. This is problematic because it prevents separating the acquisition loop from the processing loops in the acquisition pipeline. The integrated acquisition loop and processing loop makes the acquisition loop unnecessarily difficult to implement/maintain and is a direct blocker of other issues.

One approach is to eliminate multi-acquire by providing all capabilities using other acquisition techniques. One of the capabilities is to calibrate resulting data in e/eV/s (see tasklist below).

### Tasks
- [ ] https://github.com/nion-software/nionswift/issues/230
- [ ] https://github.com/nion-software/nionswift/issues/1029
- [ ] Operations resulting in multiple line plots with the same axis units should go into a composite line plot
- [ ] All 4D stem acquisitions should be able to produce 1D reductions (virtual detectors)
- [ ] Measure shift from 2d outputs (SI reduced to 2d) or possibly 1d outputs, followed by apply shift processing
- [ ] https://github.com/nion-software/nionswift-instrumentation-kit/issues/178
- [ ] https://github.com/nion-software/nionswift/issues/1019
- [ ] https://github.com/nion-software/nionswift-instrumentation-kit/issues/179
- [ ] https://github.com/nion-software/nionswift-instrumentation-kit/issues/180

[cmeyer]

Comments from email Jan 17 2024.

• the ability to automatically display composite line plot (multiEELSSpectra) could be added first as a manual post-acquisition step initially; then automatically when UI for post-acquisition steps is added
• virtual detectors are available in the "Spectrum Imaging / 4d Scan Acquisition" panel, we would lose the ability (temporarily) to do multiple sequences of SI's automatically with virtual detectors
• multiple sequences of SI's could still be done without virtual detectors using the "Acquisition Panel" right now
• "auto dark subtraction" feature probably is not functional right now; probably isn't used; and would be dropped until someone complains
• we would lose the ability (temporarily) to do automatic summing; Andreas contends this isn't useful without alignment; could still be done with a post-processing manual step
• apply shift to each "frame" (actually applied to each step in the mult-acquire sequence I think) is still available as 1D ramp; lose the ability to use with virtual detectors (again, temporarily)