I think confusion in this area often stems from misunderstanding what lexical accent is. "Stress accent" (in the case of English) and "pitch accent" (in the case of Japanese) are both types of lexical accent, a.k.a. accent on words.
Lexical accents are, primarily, used not to convey emotion, but rather to distinguish words. Each lexeme (word) has its own accent that is effectively stored with the word in your head.
When words are phonetically realized, they need to change somehow to reflect the lexical accent. In Japanese, pitch is used -- it is unusual for the amplitude to change just to reflect the accent. In other words:
In neutral renditions of an accent phrase, the accented mora only increases in pitch, not in amplitude.
However, not all utterances in normal speech are neutral.
Just because Japanese doesn't have stress accent (i.e., a lexical stress), that doesn't mean it doesn't have emphasis at all. You can emphasize words in Japanese, and that can cause changes in amplitude as well as exaggerate the lows and highs in existing pitch.
How emphasis affects amplitude and pitch
As for how emphasis will specifically affect the amplitude of each mora in the accent phrase, it varies. Here are some examples.
Sometimes the first mora gets said louder (and/or with more tension and/or with more built-up pressure):
こりゃ・たまったも＼んじゃな＼いわ
(I'm bolding the emphasized mora, not the accented mora. The accented mora are followed by ＼. So, the こ is low in pitch but is said loud, or with built-up pressure, like っこ.)
For this sentence, the beginning of the middle accent phrase could also be emphasized (which would in fact lower its pitch, and increase its amplitude):
こりゃ・たまったも＼んじゃな＼いわ
And sometimes the accented mora gets said louder:
こりゃ・たまったも＼んじゃな＼いわ
(The emphasis happens to align with the accented mora here, so the pitch of も is high, probably higher than usual, and it is said louder.)
P.S., for this sentence I think the first rendition is actually most likely.
You can also hear particles at the end of an accent phrase get higher pitch and higher amplitude, even when there is a proceeding accent kernel in the phrase:
A. 彼をですか？
B. いや、彼もですよ
This emphasis of も on the accent phrase of か＼れも would result in も being said louder, as well as the pitch going higher (for purposes of intonation, not pitch accent), resulting in an ending pitch of approximately かれもですよ{HLHLLL}.
So is amplitude completely flat in neutral utterances?
All that said, humans aren't robots, and there is no fully neutral utterance except in extremely controlled environments. In general you should expect a tiny bit of fluctation in amplitude around the accent kernel simply because you need to tense your vocal chords to generate higher pitch and you may end up increasing amplitude as a result. But the increased amplitude doesn't sound like the accent, it sounds like emphasis, especially if it's large enough. For example, in your linked audio clip, it does sound like the speaker is slightly emphasizing the word compared to a completely neutral rendition.
The important thing to understand is that if you increase the amplitude for every single accent kernel in a sentence, it will sound utterly bizarre, because that's not how pitch accent works -- almost as if you are trying to emphasize tons of random words. This is a common mistake of stress-accent-L1 learners when they study pitch accent and it sticks out considerably.
