From two pictures you can generate as many in-betweens as you want.
Will it look good? Well it depends on how different they are.
This process is more sensitive to actual motion (example caused by wind) then color shifts caused by for example a passing cloud.
So depending on the subject as a rule of thumb try not to go over 5% motion (example 1000 pixels, 5% is already 50 pixels)- and over that unless it’s a uniform motion it probably becomes a morph job.
We have a morpher too called RE:Flex which has a tool called RE:Flex morph (still morpher) where you locate some pix at point in time on the timeline and it helps you create the in-between and you help it by placing shape guides to set matching correspondances.
Of course seems easier when possible to start with more frames then less – more frames means less manual work / more controlled results.
Pierre