The bigger point is re-investment in to the scientific infrastructure - in spite of its inefficiencies feds seem to be doing an OK job but are cash constrained. On the other hand big profitable tech companies in spite of lots of cash are still playing the game of paying less for taxes and other financial engineering tricks and are doing basic minimum (which they are entitled to). One may say, its not their job, which is absolutely true. But the outcome here we want is, better investments into the scientific infrastructure so the US remains the dominant player for foreseeable future and that would not be possible with a massive companies sitting on tons of cash and who are unwilling to make investments in to the future because of the perceived "volatility".

We shouldn't be trying to shoehorn entities into jobs that their incentives don't best align with. If the premise is that we want to allocate more capital into scientific infra but the Gov't doesn't have that funding, the solution is to give government the funding, not to somehow get companies to spend their cash the way we want it. Where would this extra money come from? Again, under the premise that we want to re-allocate money towards scientific spending, it would come from higher taxes.

Fundamentally the issue you're describing is "gov't can't perform one of its competencies (scientific investment) well enough due to cash constraints". The cash constraint is a fundamentally political problem, and companies' decision-making vis-a-vis scientific investment is just a red herring.