While it's fundamental to consider that power delivered to an engine's crankshaft is not continuous, it's worthwhile to examine how impulses and their timing can affect net power. Stated another way, particularly since spark ignition timing is typically provided before pistons reach their TDC compression stroke position, not all cylinder pressure is exerted during the direction of crankshaft rotation that produces useful torque. Some of it is negative, and the idea is to reduce the time during which this condition exists. Bottom line, combustion flame rate of travel is at the core of this issue. So let's discuss some of the more salient aspects of the process.
From a practical and engine builder's perspective, among the combustion variables involved, uniformity of, and conditions for improving mixture density and air/fuel charge quality are factors that can be materially influenced to improve both on-track fuel economy and overall power. Not surprisingly, they all can be linked. For example, by creating inlet flow path characteristics and overall combustion space surfaces conducive to aiding fuel suspension, working to improve fuel atomization and helping reduce the amount of unburned fuel passing into and out of an engine's cylinders can combine to produce beneficial results.
We know that burn rates increase when air/fuel charges trend toward air-rich conditions (leaner mixtures). Since fuel-rich mixtures require more time for combustion, leaner air/fuel charges involve less time and become effectively "faster" in the process. By the addition of controlled turbulence (or other forms of mixture motion), it's possible to move toward leaner mixtures and a faster and more efficient overall combustion cycle. Of course, the smaller the combustion surface-to-volume (s-to-v) ratio, the less heat is lost to exposed surfaces and coolants. The net effect is increased heat converted into useable work at the crankshaft. Small combustion chambers and flat or near-flat piston crowns help contribute to keeping combustion s-to-v ratios low.
However, it's possible to over-do this concept if it turns out combustion volumes are sufficiently small, relative to the total combustion surface or mixture motion becomes too aggressive. The result trends toward an increase in non-participating portions of fuel that then pass directly out of the combustion space unburned, due largely to surfaces at sufficiently low temperature to prevent this fuel from burning or fuel that has mechanically separated from the air. One practical approach is as previously stated--small chambers and flat or near-flat piston crowns and non-linear flow paths that don't upset mixture homogeneity.
Dating back to the days of Sir Harry Ricardo's "High Ratio Compact" chamber and then to around the late 1970s with Michael May's "Fireball" high-turbulence design, creating specific mixture motion patterns in the presence of low combustion s-to-v ratio environments has enabled so-called "fast burn" engines to develop high specific power and efficiency levels. Circle track engine builders have long understood that by whatever means they could create similar conditions, on-track throttle response and off-the-corner torque were verifiable results. However, aside from improvements in fuel efficiency and higher cylinder pressure, we need to review how a fast-burn condition can penalize power if not accompanied by consideration of other engine characteristics, notably spark ignition timing.