Strange CompactLogix issue

Ok technology guru’s. I’ve got a very strange phenomenon happening with a CompactLogix (L35E) controller. I’m reading and writing to a device that has I/O data with four message instructions. Basically I’m reading from data table A and writing that info to data table B and reading from data table B and writing that into to data table A. There are two versions of the logic (the two attached files). In program 001 when I read the data from one table I’m putting it directly into the tag that is then written to the other table and vice versa. When that program is loaded and I power up the PAC it take 45 seconds for the processor to boot up and when it does I get immediate results on the I/O data. In other words as soon as the processor boots I see the data being moved from one table to another. However, in the other program, 002, instead of putting the read data into the tag of the written data I put it into a tag and copy that tags data to the tag that is written back to the device. I need to do this because for the final version of the project I need to manipulate the data. The problem is that when I do this it takes an additional ten seconds after the PAC boots up for the data to start moving. The messaging starts as soon the PAC boots (I can see it in wireshark) but the data isn’t being updated for just under ten seconds every time I power up.
I’m not by any means a programming expert and there is a lot I don’t know how to do so I’m guessing I’m missing something but I have no idea where to look. I’ve tried all kinds of “tricks” in the logic but the results are always the same.
Anyone have any ideas?

I forgot one last thing. If when I power up the PAC I have it in program mode (via the key switch) and switch it to run at the 45 second mark the I/O starts responding (as opposed to it taking ten seconds)

Well I’m not sure anyone will care but I figured out a fix. I have no idea why it works but it does. I added a top rung with a “f:sf” examine on bit that latches and unlatches a couple of bits. It’s putting them back to where they should be when the program starts and when the PAC fully boots the I/O starts responding instantly. Like I said I don’t understand why this works because the bits have nothing to do with the tags that move the data around but it’s working.