@MeSophie this is impressive. Perhaps @lilyclements may have suggestions to improve on the new layout of the controls. But it is better than I could have done.
a) I first checked that the original definition is still working. Pretty good, but I think I have found one problem. It may have been there before.

1. I tried with dodoma and 20mm in 3 days from 1 Nov to 31 december. Saved all 3 variables and that worked fine! Great!
2. Now I added your additional option, of the default dry spell condition. Named them differently, so we have both - now 7 variables in the summary. Seems fine - great.
3. Now unticked the Additional, so I'm back to the original, i.e. situation 1. Renamed again to get a total of 10 columns. Now I found I got the same results as in 2), i.e. still with the dry-spell condition. It should have taken that condition away, so I should have got the same as 1).
4. Now I did a reset and tried again. It now gave me an error, that mentioned dry spells. So I wonder if it is not clearing properly? I then went into the sub-dialog, and saw that it had cleared the dialog. When I now tried again it all worked!

Now I moved to check your new evaporation option. Here is the dialog I used:

And, to my amazement it worked first time. Congratulations to @lilyclements and @MeSophie. Just one small thing - of course! Here are the results:

Now I had made the evaporation column using the calculator, to be simply 20/3 . So, with the fraction as 1 I should get the same result as before. And I usually do, but not always! I wonder why not? So I had a look at the data, for 1951 when there was a big difference. I was happy with day 350 for the ordinary start, but the evap had started a lot earlier, when there was one day with 14.7mm. So I think there needs to be a small tweak in the calculation system for the evap situation.

Here's the data in 1951. I am enjoying using the pivot tables to look at the climatic data:

How exciting to be so close so quickly though!

@rdstern I was unable to reproduce this error. Everything works fine at home. I don't know if Lily will be able to reproduce it.

@lilyclements Could you please have a look at this part when you have time? Thank you.

@MeSophie @jkmusyoka has just found an odd error in the start of the rains. It may (or may not) be due to using the shifted year - I started with Zambia data from August.
I haven't seen it before, and perhaps it is because I went straight to the start of the rains dialog. However, I wondered if it would be just the first time I used the dialog, but I then got it whenever I repeated, even after the data frame was created earlier.

It is annoying, but it doesn't seem to stop the results being created. Perhaps you could investigate as you continue to improve this dialog. By the way I got it with a simple definition of the start, and nothing to do with evaporation.

Thanks

@rdstern @jkmusyoka We have found a solution for the second error reported in #9185. Please have a look.
We were unable to reproduce the first error.

@MeSophie great that you can include the solution to @jkmusyoka error. Can you also check with @lilyclements on the problem above, that the current R code seems to be incorrect in the date of the start occasionally. And @lilyclements may also have some ideas on further improvements in the new layout?