Courtesy of David Kamin, I note some additional issues that the use of scrip might raise. Both as a legal matter and in practical terms (especially for people who are cash-constrained), scrip is not quite the same as cash. Thus, while offering it is probably better than just not paying people, its use would still pose such problems as the following:
--Paying off bonds with scrip would still technically be default. So cash would still be used for that, and also perhaps for various other purposes. So the Treasury would still be prioritizing payments, raising both political and legal issues as well as that of technical feasibility (the computer systems problem).
--It's up to the mysterious wisdom of the bond markets, even if bondholders still get cash, to what extent the use of scrip to pay amounts that were legally due in cash avoids weakening our reputation as a reliable borrower.
--People who are relatively poor and/or cash-constrained would especially face the non-cash equivalence problem. Beyond merely earning zero interest if they just hold onto it (which of course is equally true of cash), they might be hit with a steep discount if they tried to sell or deposit it. This would not merely raise distributional concerns, but might also have a macroeconomic impact on consumer spending - again, if one compares it to finding a way for the government to keep paying out cash, rather than to just doing nothing.
In sum, I certainly agree that using scrip is both (a) likely to be much better than just doing nothing and (b) worth considering as perhaps the ultimate best course to follow. But given these downsides, it is not necessarily the best option. I myself still tend to favor the platinum coin, although ultimately this is an empirical question (turning on how the various alternatives would in fact play out, economically and politically) to which I cannot claim to know the answer.