Default filtering table with null and zero values: when click "0" it lists all 0's AND null items

> When you filter for the value 0, the Blanks are cast/treated as 0 so you see all rows in the table that are either 0 or Blank.

Yes that's the whole issue, I found an older post where someone stated it's by design and won't be changing.

I am now using COUNTROWS instead of the built in count summarization and this properly counts blanks separate from 0's, but the issue stated in this question is still there, one visualization shows that I have two 0's, yet when I click that row to see all the rows with zeroes the other table shows zeroes and blanks. There is no solving this, there is only working around it because the underlying implementation for cross filtering uses different rules than how the table shows counts, and Microsoft has said the underlying implementation is by design.

To continue my rant, I will state again that one visualization shows one set of data, and filtering on that data results in a different set of data on other visualizations. Blank/null is not the same as 0. As programmers who use nullable types, as Power BI claims to support, we are used to setting things to null separate from setting things to 0. For example, if I'm reading temperature data every second for an entire day using a probe that only produces integer outputs due to it only being accurate to +/- one degree, I would code my program such that it produces "0" when the temperature is 0 degrees, and would produce "null" when the probe was not working for some reason. Power BI does not let me see when all the times temperature was 0, it simply does not work when there's null data. I have to manually filter using workarounds when I have 86,400 data points for a 24 hour period and faulty probes producing 10,000 nulls every day. 0 does not mean false for temperature data, I am not going to change the datatypes to handle this issue. If Power BI claims to support nulls then I am going to use them. If Power BI did NOT support nullable integer types, then yes I'd be forced to use sentinel values, "a temperature of -300 degrees was when the probe wasn't working", now all of my bar charts have to be tweaked because they extend all the way down to -300 degrees, yet when there's null data they properly just ignore the null data and there's zero extra work for me.

Manual filtering is a workaround, the underlying issue exists whether I do a separate page only filter to hide all the blanks or not. I have to give end users long winded explanations on why there's a table that shows separate counts for null data from 0 data but that they cannot do the natural thing of clicking on the 0 row to get a list of all the 0's. "Oh if you want to see the 0 data you have to get rid of all the blank data first via a separate page filter. Yes I know the dashboard works one way in all other cases, but it sometimes equates 0's and blanks (not when it's counting, though!), so here's all the instructions on how to work around that".

The simplest clarification Microsoft could make is for the filter to highlight both the 0 and the null rows when you go to filter on 0, then the end user can see, "oh I can't filter on just 0, I can see that because when I click the 0 row both the 0 and null rows are highlighted". Instead we get false information, the 0 row highlights and the null row stays grayed out, so one thinks they've now filtered to all the 0's and not the blanks, but that is misinformation.