The discussion on an earlier post to this community discusses the resource needs of trying to compile this data on the fly and of course this simply comes down to use of function vs. revenue to supply the equipment to do the task.
While I don't have hard and fast data (though I'm willing to try to write some analytical software to supply said data), I would estimate that at least 1/2 of the 'active' users (interpret active as you will) are sitting up against this 150 interest ceiling. This suggests there is at least demand to increase that limit. Doing so costs processor time. So, this would seem the perfect place to allow users to push a revenue stream into LJ. It seems to have worked well with user icons which is another purchase to expand function. Whereas with interests, the cost is more storage space that computational.
I'd be interested to hear some commentary on the possibliity, likelihood, and even the limitations that this idea has. I do not know a great deal of the internals of LiveJournal and can only many educated guesses as to the specifics of the resource needs for this.