2F/2F has provided a very good explanation of the considerations. One potential difficulty using the Rollei material is that it doesn't really extend much into the infrared range. Last summer I came up with incident metering at 400 and increasing exposures 6 or 7 stops with a 720 filter or about 12 or 13 stops(!) with a 760 filter. The latter gives more IR effect, but you are really working down the cutoff slope of the film. With the EFKE IR820, there was only a stop or two difference between the filters.
At least one discussion last year came up with a problem that suggested a guy had a miss-marked filter also. You might be able to tell a K-1 from a K-2 by looking at it, but 750 versus 850 nM IR, not likely.
I still think a lot of normally intuitive stuff goes out the window because we don't see the spectrum the film sees. Where I might normally bump up a half stop or a stop to "improve shadow detail" in a visible light shot, I have no idea that oak tree looks like a giant cotton candy in IR, for example.