Thank you for all the votes, responses, and private messages on this so far.
The data points I am working from are the Lycoming IO-390 manual and the Vaughn Askue book "Flight Testing Homebuilt Aircraft," which I purchased from Van's Aircraft.
The engine manual specifies two test runs on a new engine. The first is a ground run:
* 3 minutes at 1,000 RPM
* 15 minutes at 1,500 RPM
* Check magnetos
* Cycle propeller
* 10 seconds at "full static power"
The second is a flight run:
* Full power takeoff
* Reduce to climb power as soon as possible (25" / 2500 for the RV-14)
* Shallow climb to ideally 5,000 feet, no more than 8,000 feet DA
* 1 hour at 75% power (26.3" / 2350)
* 1 hour alternating between 75% and 65% power (23" / 2350)
* 30 minutes at maximum continuous power
* Descend at 65% power
By contrast, the Askue book strongly advises doing land-backs, a variety of flight speeds, and some stalls during the first flight. Those are the items in conflict with the engine manual.
I took the RV-14 transition course and I have stayed current and proficient in a variety of planes, including a few different taildraggers, so I feel comfortable with my ability to recognize abnormal behavior in the RV-14 as well as to safely get the plane on the ground and stopped from any reasonably normal behavior it exhibits in the air. But obviously I want to minimize the risk of being in the air in an uncontrollable airplane to the maximum extent I can without compromising the health of the engine.