Part of me can't help but suspect that they are also doing this because, believe it or not, many computer science researchers don't know that these paywalls exist, because they are subscribed through their university. Now that they are stuck at home, these researchers could start noticing how inconvenient and ridiculous these paywalls are, given that they are the ones who write and typeset the papers and review the research for free.
So it may be a smart move for ACM to avoid making researchers aware of paywalls, and continuing to drag this outdated publication model for a few more years... (If I'm not mistaken, most subscription money comes from universities, which subscribe annually, and there's very little revenue from individual downloads, so most likely this measure isn't actually costing ACM much in terms of revenue.)