Per the original question. A respected engine builder once said:
For a given engine combo if I have a known good cam profile I add 8 degrees of duration @ .050" lift for each additional 50 cubic inches. I also like to tighten the LSA 1 to 2 degrees for each additional 50 inches.
Both combos will HP peak in the same RPM window and have similar idle characteristics...
He was talking about building a similar but larger engine from the same engine family with the same or similar parts. Just a displacement increase.
I assumed he tightened the LSA because the larger displacement motor can tolerate more overlap (in addition to the increased overlap from more duation) and still have similar idle and low rpm characteristics. And also because the larger displacement motor needs even more overlap to help breathing at high rpm, given that it has the same or similar induction system trying to feed more cubes. I'm not sure, but that's what I assumed.
The 8 degrees per 50 cubes would yield 14.4 more degrees for a 90 cube increase which seems to be in line with the other replies.
The big variable not addressed by this, which Ross mentioned, is the rest of the induction system. I would think if the heads, intake, etc. were marginal on the small cube motor, then the large cube motor would need even more cam duration added to keep a similar rpm range, and the results would not be very good. If the heads, intake, etc. were replaced with higher flowing pieces on the bigger motor then you might need less cam duration increase than the formula calls for. The rpm range is determined by displacement and the flow characteristics of the entire induction system. The cam is just one part of that. That's the thing with rules of thumb. They by intention oversimplify the situation. They can still be useful though.
JMO,
paulie