I have figured out many of the settings but some still confuses me (
Bits per pixel: should I use 16, 24 or 48 bit RGB?
Scan resolution: I set it to 1200 (max). Appropriate?
You should scan at 16-bits/channel, so 16-bit gray for Grayscale images, and 48-bits for 3 channel RGB images. Whether you want to also output in 16-b/ch after gamma adjustment depends on how much postprocessing you intend to do. If you want to do a lot then stay in 16-b/ch, otherwise you can create smaller output file sizes by then dropping to 8-b/ch (8-bit Gray or 24-bit RGB color).
Depending on your original images, there may be little resolution in your prints beyond 600 PPI. Only you can tell. You could scan at 1200PPI (I presume the scanner's native sensor resolution), and downsample when producing the output.
White point (%): 0.1. I read somewhere that this would be a better than default (1.0). Opinions?
Just look at the white areas of your images, if you want to retain a little paper structure there, 0.1 is probably better. You can see in the status bar at the bottom what RGB values the white areas have, when they are a tiny bit below 255,255,255 then you have retained some tonality there instead of clipped white.
Scanner color space: Built in. I don't know how to make a profile and not sure if I need to...?
If you don't have a dedicated profile for your scanner, just leave it set to "built-in".
Printer color space: ProPhoto color space
No, this should reflect the profile your printer requires to make good output. So it is either a specific profile for a specific paper type on your specific printer, or just use sRGB. This choice is only used when you print directly from VueScan, e.g. when using your scanner as a copier.
Output color space: ProPhoto color space
This should be set to the working space of your photoeditor application. Programs like Photoshop can use a colormanaged workflow, with conversions between colorspaces. When the original contains hugely saturated colors, then "ProPhoto RGB" can be beneficial, but when your originals are prints, then "AdobeRGB" is probably more than enough. If you don't have a colormanaged photoediting workflow, the set the output color space to "sRGB".
TIFF size reduction: 2
This downsamples the output image by combining/binning pixels. You could use that if you initially scanned at 1200 PPI but want to reduce the file size to 1/4th (half horizontal and half the vertical number of piixels).
TIFF file type: Again, 24 or 48?
If you want to do more postprocessing on the files, keep it in 48-bit mode (3x16-bit channels) till you're done, and if not then you can reduce the file size to half by using 24-bit (3x8-bit channels).
TIFF compression: Auto
What's the difference between reduction and compression. Do I reduce quality if I reduce with 2 and have auto compression.
See above, reduction removes pixels by averaging blocks of pixels before reducing the file size to single averaged pixels. Compression is chosen to make the files smaller but without losing any pixels. Leave this to automatic, it will make the best choice for you.