I didn't want to consume my own time trying to find it in related documents, as you have. Plus, if the OP didn't use the 1k the OP probably hurt the chip in which case nothing will work. I was assuming that if the OP was programming the OP would have used the 1k, but I couldn't find it in the main datasheet. So, the OP probably did not use the 1K which might be another problem. However, EA and PSEN must be in a known state prior to anything happening with the RESET logic, regardless. This just deals with consistency and keeping everything stable as with good housekeeping practices. I mean one can have odd circumstances where the chip just does not recognize the response. With the OP saying that he has done something with the RESET pin and then has done something with PSEN is just the wrong sequence, even if one assumes that PSEN and EA can be messed with during the time before the falling edge of a reset signal. I have to say that I do not like how much trouble one has to go through to get these chips to program. It's nice when you can use the firmware to auto-jump to the bootloader via a command, but going through hardware conditions is just too finicky. Be nice if they used one of those NC pins to have logic incorporated for the user to reduce the ISP condition to a single pulse.

Also your document which is on the page that I linked with earlier states


But the datasheet makes no mention of the 1k. This is just misleading. Also, in that document they state that the ALE must be left floating. I would just try the safest combinations and see if I couldn't get it to program because there is too much ambiguity.