Not really IMO. Creating more frames only to remove them later (which is what will happen when you adjust the time in a video editor, assuming you’re using the same frame rate in all instances) seems both redundant and a possible source for glitches – how good is your video 'ware at re-sampling? Will it re-sample at a perfectly consistent rate? If not, you may get less-than-smooth results.
What seems the better way to go is to determine the exact length (i.e., time) you need for the rotation sequence, how much rotation is best in that time to get the smoothness you want, then render just the necessary number of frames (with perhaps a little head and tail padding) and no more.
Finding the right timing may take a few trial runs, but here’s a way I’ve used that cuts down on total Blender rendering time:
After making your best guess at the sequence time and rotation rate, rather than doing a full render, output a “playblast” using the “Render this window…” button that’s at the right side of every 3D Window menu bar. CTRL+RMB this button to do a quickly-rendered animation of the 3D window view – no lighting, no textures, just the movement, but that’s the important thing at this point. Use whatever your target frame rate will be. You can output to any convenient format just as with a full render.
Place this “playblast/pre-viz” in your video editor and try the timing. If it needs tweaking, you can adjust it in Blender and run out another quick-render. Once you get it as you want it, set up for full render and make the “beauty” version.
In the past I’ve used a video app (Virtual Dub) to output versions of a base sequence with different lengths by jiggling the frame rates on output, but I’ve found that the playblast approach is both faster and more accurate; you can get things perfect down to the exact frame before spending the time on a full render.