When to use execution plots
As others have said, kill-plots are now fundamentally different from assassinations. The choice of method is much more dependant on circumstances. Considerations, based on one night's (er, morning's) play:
1. How urgent is the kill needed? Assassinations will always be quicker, while plots can be rather random. Higher plot power certainly reduces the amount of time taken to gain a kill, but (as with all randoms) does not guarantee anything. If the target absolutely needs to die, right now, pay for an assassin. If the target's death is one of a range of possible ways of usurping their holdings (or whatever), plot.
2. How vulnerable is the target vs their title? Assassinating a high-ranking target (for example a Duke) costs a lot of money simply because they have the title. But if the target is a child (especially if they are a child) with little support among their own courtiers, plots can be an extremely cheap and effective method of removing the target.
3. How involved are you in the politics of the target's provence? If you don't care much about the politics of the target's provence, and just want them dead, assassinate. If you are already deeply embroiled, with many friends in high places, plots become a logical extension of that involvement.
4. Is the target's Spymaster or partner supportive of your plot? Plotters need to be trusted by the target - merely having a lot of plotters is immaterial because certain events appear only to trigger for individuals in trusted roles. I can't yet be more precise, but at this stage note that there very large differences in the effectiveness of individual potential plotters.
5. How highly do you value the plot slot? Sometimes plots become essential mechanics in a game-plan, and need to be reserved for plots that take a finite time to complete (for example, revoking a title). Perhaps your large nation can't afford to have the plot slot filled with a 3-year long mission, but one can afford an assassination. Other times plots are marginal, essentially free uses of slot that otherwise isn't being used for anything. These plots tend to be more useful for smaller nations.
I can't yet vouch for how well balanced these new plots are. Inevitably there will be some frustrating experiences at the start, even if one understands the strategies above: Plots are still random, and sometimes randoms are too random. But I like the design intent behind them.