Clarify that CPU is between samples and category&stack are at sample time

When CPU numbers are shown, the tooltip looks like this:
So it's something like this:
> Category: **[] Layout...**
> CPU Usage: **97% (...)**
> Stack:
> **parse_into()...**
> **...**

I'm concerned that we're mixing two "domains" of information, and it may give the wrong impression, e.g.: "Wow, this parse_into in layout is taking 97% of a CPU!".
But in fact, all we can say from this data is: Since the previous sample and now, we have used 97% of a CPU; And at this exact sample time we were calling parse_into in layout.

So I think we should clarify the difference.
Suggestion:
> CPU Usage since previous sample: **97% (...)**
> .....................................................................................
> Sample at time: **0.8465s**
> Category: **[] Layout...**
> Stack:
> **parse_into()...**
> **...**

This clearly separates the more continuous CPU numbers from the point-like sample data.

As a bonus it shows the precise timestamp. I don't know how valuable that is, but for example it could help find markers at that time in the Marker Table.

┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FP-372)