I have found a study [1] of 2017 (http://www.sharetheluck.ch/single-post/How-rare-are-four-leaf-clovers-really) which answers a great part my question, and I'd like to show those results together with some additional evaluation.
They have counted 5.7 million clovers and found the following numbers compared to 3-leaflet clover:
4-leaflet clover 1 in 5,076
5-leaflet clover 1 in 24,390
6-leaflet clover 1 in 312,500
For the 4-leaflet clover they have corrected appreciably the common figure of 1 in 10,000 (see e.g. the comment of Untitpoi here).
A logartihmic plot of these relative frequencies shows interesting features
(i) a fairly regular exponential decay of the frequency in the range n=4,5,6
(ii) a significant "exception" from this rule is the "normal" clover with n=3

How can these observations be explained?
It would also be interesting to extend the study to n>6, and maybe also to the range of "deficient" clover which posesses only one or two leaflets. The latter are most probably very rare so that the complete log-plot would resemble an inverted letter V.
The question of 2-leaflet clover was raised here [2] https://www.quora.com/What-are-the-chances-of-me-finding-a-two-leafed-clover. But no quantitative answer was given.
References
[1] http://www.sharetheluck.ch/single-post/How-rare-are-four-leaf-clovers-really
[2] https://www.quora.com/What-are-the-chances-of-me-finding-a-two-leafed-clover
[3] https://en.wikipedia.org/wiki/Four-leaf_clover#cite_note-auto-6
A comprehensive reference which I have only partially explored up to now