That's an extension of the "bookkeeping" version of Say's law which is, as I outlined above, obviously correct.

You are referring to the version of Say's law that says that all production WILL be bought, and that economy will always work at maximum output.  THAT version is of course wrong, because it goes one step further.

The "bookkeeping" version of Say's law tells us that all production CAN be bought.  However, a stronger version (in fact, the original version) of Say's law states that it WILL be bought.  Now, *that* is not necessarily true.  You might very well decide, even though you have the money, NOT to buy the production that remains to be bought, simply because you're not INTERESTED in buying it.  In other words, one can produce goods which are not in demand.  The earnings to buy them are available, but nobody wants to buy them.

I use Say's law in its more restricted version, which simply states that it is impossible to have, *by lack of earnings*, overproduction that CANNOT be bought.  I'm not using Say's original version of the law, that deduces (erroneously) from this, that it WILL be bought.

So, yes, I agree with Keynes' rejection of the strong version of it, exactly because of lack of demand.  But that was not what was discussed here.  I mentioned (the restricted version of) Say's law to contradict the statement that if wages are not high enough, there is simply not enough EARNINGS to be ABLE to buy the production. That is obviously wrong.