Whoa that is a lot of lines!
I've been out of the field for several years, but before TBC added the ability to key in a level file, I used to code these M5.dat files by hand (for importing differential-trig runs using a total station), so I have done may fair share of troubleshooting on them.
Just make sure you know, the import summary is very useful for M5.dat files. It gives you the address line where there is a problem. That helps you focus your troubleshooting. Out of sequence errors can be ignored, and usually mean the file was edited, but not in this case.
I think all of those errors are associated with a cancelled observation, which is the last one in the file (somewhat luckily). Here is the relevant section:
For M5|Adr 2280|TO Start-Line BF 61| | | |
For M5|Adr 2281|KD1 BM2047 61| | |Z 437.320 ft |
For M5|Adr 2282|KD1 BM2047 3 61|Rb 7.804 ft |HD 292.85 ft | |
For M5|Adr 2283|KD1 127 3 61|Rf 3.467 ft |HD 242.75 ft | |
For M5|Adr 2284|KD1 127 61| | |Z 441.658 ft |
For M5|Adr 2285|KD1 127 3 61|Rb 7.545 ft |HD 254.00 ft | |
For M5|Adr 2286|KD1 ##### 3 61|Rf 3.722 ft |HD 245.83 ft | |
For M5|Adr 2287|KD1 ##### 61| | |Z 445.480 ft |
For M5|Adr 2288|TO Measurement repeated 61| | | |
For M5|Adr 2289|KD1 128 3 61|Rf 3.723 ft |HD 245.93 ft | |
For M5|Adr 2290|KD1 128 61| | |Z 445.479 ft |
As you probably noticed, this run isn't being shown on the import editor. I would take this run out of the file and see what the import looks like with it removed. I suspect that it will come in fine.They pick up on BM2047 (Adr 2281 & 2282), foresite 127 (Adr 2283 & 2284), and backsite 127 (Adr 2285).
After that it appears that they went to forsight something and cancelled it. Then they foresighted 128 (Adr 2289 and 2290) and off they went. The parsing error is associated with Adr 2286 &2287. I think it doesn't know how to handle the lines.
Just looking at the math, I think you'll be fine if you remove the entire lines for Adr 2286, 2287 and 2288. You will still see an error in import that says line 2286 out of sequence. You can disregard that error because that one doesn't cause any issues for importing the data. You can renumber all of the following ones if you want, but you don't have to (I often did when I was feeling particularly on the spectrum). If you get an error that tells you the file's math doesn't 'add up', then there is a bigger issue, but I don't think you will here.
Finally, because the overall level work doesn't appear progress linearly, I think may be best to process this with Least Squares. Make your life easier by eliminating the import of temporary (not to be used again) turn points. TBC counts the number of turns so it will propagate the error correctly. Then I always added a geographic coordinate to all points I didn't have a good Hz position for. This allowed me to view the level file in 2D (another one of those things that is spectrum related, lol).
Unsolicited advice:
I'd recommend that your crews enable curvature and refraction correction. Having well balanced turns and runs can mitigate this to a large degree, but that wasn't always done in this survey. It is simple, reliable and effective. In my 30 years in the field, I always used balanced turns. This includes using yard rods (3 wire), which had procedures that mandated it. When we started using trig-differential leveling we continued to balance turns, which is necessary. When the electronic levels came out and enabled 3 wire level of precision and automatic curvature and refraction correction, I still kept the same balanced-turn field procedures. I am a surveyor with a large state Department of Transportation in the USA, by the way.