Now you could present me with the statistics in how many months BRK did move more than 20% (or my 15%) ... For me this means: Selling covered calls only(!) when Price to Book Value is 1.5 or higher!

Darn, I closed my spreadsheet without saving it, so I have to do it all over again. No biggie.

For all rolling 6 month periods, 32.5% of them BRK gained more than 15%.
23.7% of them BRK gained more than 20%

For all rolling 12 month periods, 49.0% of them BRK gained more than 15%.
41.1% of them BRK gained more than 20%

Monthly, only 3.0% of months gained more than 15%. And 1.7% more than 20%.
The most recent was Feb 2000 which gained 30.0%. But Jan 2000 lost -14%. In the 4 months before Feb'00 BRK lost -23%. In fact, 9 of the preceding 12 months had a loss.

I do not have historical P/B numbers. If anybody has that data and can email it to me, then I could incorporate into the spreadsheet (which I did save this time).

"In down a lot, you fully participate in the loss. This is infrequent, but gives you a large loss."
We are talking about the strategy of LTBH Berkshire shareholders to make a few extra bucks, people who hold Berkshire through thick and thin and wouldn't sell in this case anyway, whether they wrote covered calls or not.

Now you are mixing two different things. Either you recognize the loss or you close your eyes and ignore the loss. The loss is there whether or not you chose it recognize it.

The point is that you chop off the gains at the strike, thus missing out on the gains above the strike, but you fully participate in the losses.
Yes, you always keep the premium.