Warren was once quite comfortable in this field. Why not so much anymore?

Because real bond yields are low.

Considering that:
(a) what we call "cash" is only a shorthand for the short end of the entire fixed income portfolio, and that
(b) the fixed income portfolio (including cash) isn't any bigger than historically normal as a fraction of the assets of the company or as a fraction,

...the conclusion is that the cash pile isn't particularly big, and won't be even at $200 billion, because there is almost nothing in the way of long term paper.
As the cash pile isn't a particularly big allocation, there is no reason to bemoan the fact that it hasn't yet been put to work or search for reasons that it hasn't been.

The only thing that happened is that the fixed income portfolio has been moved firmly towards the short end of the curve. Fewer holdings with durations over 12 months, more with durations 3-12 months. This is the reason I give the answer "because bond yields are low" above.

Cash + fixed income as percent of investments per share at end Q1 = 36%.
It was a that high or higher at every single year end from 1999 to 2019 inclusive, and averaged 48% in that stretch. We are 1/4 underweight fixed income (including "cash"), in that sense.
And the flip side: listed equities as a percent of the investments per share is much higher than the old average number (post Gen Re): much more money is currently at work in the markets than the historical norm, not less.

It's easy to forget how much bigger the company is now. There was a time that $150bn would have been an outrageous pile, but no longer.

Jim