I don't know that it's risk free. There is risk to keeping payment channels open, there's the risk that your node goes offline and the counterparty closes the channel. Even with that risk though I do think LN BTC is worth more than on chain BTC, but only up to a point: the value difference between LN BTC and on chain can only go as high as the fee disparity, and probably not that far due to the risk I mentioned above. And ultimately, channels have to be closed eventually, which can be looked at in 2 ways: that either that added third fee for closing the channel makes BTC on an unclosed channel that much more valuable, or on the other hand that ultimately for finality the LN BTC has to price in that cost and therefore is worth *less* than on chain BTC that has already borne that cost. My write up you referenced talks about how ultimately fees break the store of value use case and collapses the value of BTC, and I think that LN BTC wouldn't fare well in that scenario because ultimately it's only secure because it finalizes on the blockchaim eventually. All in all, any arbitrage opportunity requires those 3 fees to be paid (the two you outlined, moving from exchange to wallet, moving from on chain to a payment channel and the one I mentioned, the impending cost to close it) and so those fees would eat up any arbitrage opportunity. The only real way for there to be one would be if the privacy and speed of LN alone justified the price premium.

As far as the political side of the article, political instability does affect perceived value of fiat currency, this is empirical, I don't foresee the supreme court going the way the article speculates.