Marking a batch of photos one at a time is tedious and produces inconsistent results. Doing them together fixes both problems.
A small mark in the corner is easy to remove with a crop, which makes it decorative rather than protective. If the point is to deter reuse, the mark needs to sit somewhere that cannot be cut away without ruining the picture.
There is a real trade here and it is worth being deliberate about it. A mark large enough to be useful is also large enough to be intrusive, and a mark subtle enough to be invisible protects nothing. Decide which you want rather than splitting the difference badly.
Use a transparent PNG. A logo on a white square puts a white box on every photograph, which looks like a mistake because it is one.
The size of the logo is set as a share of the shorter edge of each photo, so a whole batch comes out looking alike even when the photos are different sizes. The default of 7 percent is small and tidy, and raising it makes the mark more prominent.
Try it first on one landscape and one portrait photo, because a position that suits one can crowd the other. A logo as a PNG with a transparent background looks best, and each photo is saved separately with marked added to its name.
Yes, text works the same way and is often more legible at small sizes.
For photos of the same shape, yes. Mixed portrait and landscape sets are worth checking individually.
No.
Around 40 to 60 percent is a good balance: visible enough to credit you, faint enough that the photo is still pleasant to look at.