Estimated reproduction numbers for #Covid19 in Switzerland seemed to imply a small effect of the lockdown. (see e.g. https://www.tagesanzeiger.ch/ansteckungsraten-flachten-bereits-vor-dem-lockdown-ab-809893127675)

A preprint offers a possible explanation.

"We argue that the method used to derive the [curve on https://ncs-tf.ch/en/situation-report] is systematically flawed and leads to an underestimation of the efficacy of the lockdown."

https://www.medrxiv.org/content/10.1101/2020.05.12.20099366v2.full.pdf

This would be due to the noise, overshadowing the real effect (see R_fit vs R').

Note that the authors of the R estimations R write: "We improved the method to infer infection dates from the observed data by using a deconvolution step as discussed in two recent pre-prints" which include the one preprint mentioned in this thread

https://ncs-tf.ch/en/situation-report