use delay_lp in estimate_truncation

[sbfnk]

Closes #348.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if e1121a1 is merged into main:

•   :ballot_box_with_check:default: 47.2s -> 47.7s [-17.59%, +19.7%]
•   :ballot_box_with_check:no_delays: 50.4s -> 46.4s [-16.06%, +0.05%]
•   :ballot_box_with_check:random_walk: 15s -> 14.4s [-15.96%, +8.59%]
•   :ballot_box_with_check:stationary: 28.9s -> 30s [-4.68%, +12.24%]
•   :ballot_box_with_check:uncertain: 1.05m -> 1.08m [-4.98%, +11.17%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[seabbs]

Nice - this is great. Someone asked me today about other dists for estimate_truncation so definitely a use case. Also will help clean up the code once 2.0.0 is out and all the depreciation handling has been removed.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 34277d2 is merged into main:

•   :ballot_box_with_check:default: 46.8s -> 42.6s [-26.48%, +8.53%]
•   :ballot_box_with_check:no_delays: 50.6s -> 45.1s [-22.14%, +0.45%]
•   :ballot_box_with_check:random_walk: 14.3s -> 13.3s [-17.18%, +3.21%]
•   :ballot_box_with_check:stationary: 28.9s -> 28.8s [-10.5%, +10.04%]
•   :ballot_box_with_check:uncertain: 1.08m -> 1.08m [-8.61%, +8.42%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[kgostic]

Wow, thanks guys! We are going to use this immediately :)