I think the answer is no one, but the typical navigation text and reference publication would not be the place to learn that.
Look at these sample definitions from the standard reference for chart reading called Chart No. 1:
There is no further clarifications of the terms in the booklet. There is definitely an implication here that we need to know what these chart datums are, but we do not. This type of presentation is just another way the government supports navigation schools.
In the top picture the phrase
"Drying heights and contours above chart datum"
can and should be replaced by
"Drying heights and contours above the water level when the tide height is 0."
"MLLW (Chart Datum)"
should be replaced with
In fact, there is no place in practical navigation where the navigator needs to know what tide level is being used for the chart datum. The depths printed on the chart (called soundings, because the numbers shown on the chart were actually measured at some point) are the expected depths of the water when the tide at that time and place is 0. If the chart reads 13 ft, and the tide is 4 ft at the moment, then we should expect the water to be 17 ft deep at that time.
Put another way, we should always expect the water depth to be deeper than shown on the chart except in those short periods of time when there happens to be a negative tide.
All nautical charts of all nations work this way.
The chart datum used is actually some arbitrary level of the tide averaged over a long period, usually 19 years. On the US west coast where we have two highs and two lows each day and the two lows are not the same, NOAA uses the 19 year average of the water depth at the lower low of each day and they call that the depth of zero tide, labeled MLLW. The idea is get some reference plane that will make the tide positive most of the time.
They could have done something more like the Canadians and decided not to use the lower lows every day, but just the lower lows on the days near the new moon and full moon. This would lead to even fewer days with negative tides.
It really does not matter what they do, so long as the folks making the charts and the folks making the tide tables are still friends and working together. Beyond that the navigator really does not care.
By the way, with all that said, I must add that this lack of concern does not apply to the vertical height datum used on charts. That is always Mean High Water (MHW) and it is crucial that we know and understand this concept. In fact, the nautical chart is the only place to learn the value of MHW. This crucial data (ie for predicting bridge clearances) is not available in tide tables. It can also vary from one part of the chart to another, in which case there would be multiple listings on the chart..