I wouldn't say nearly useless. For my "most cited" paper, the vast bulk of the citations from said paper took place several years after it went into open acces

I wouldn't say nearly useless. For my "most cited" paper, the vast bulk of the citations from said paper took place several years after it went into open access because it was NIH funded. Heck, "peak citation" was this year, a full 8 years after it was published. The same general trend is true for my second most cited paper - peak citation is several years after the fact. Both of these papers were also trivially easy to find in PDF form online.