Using WinHTTrack Website Copier 3.49-2

Especially with large downloads (gB's) I encounter all the time all sort of
problems (not download-errors) which makes the use of WinHTTrack really
problematic, time consuming and untrustable.

Problems I really don't understand why.

As an example :
On a site there are plenty of files, lets say with the extentions abc, def,
ghi a.s.o.

Example:
File1.abc 
File2.def
File3.ghi
.....

A lot of those files are downloaded first as

File1.html.tmp
File2.html.tmp
File3.html.tmp
.....

and become in a later stage

File1.html
File2.html
File3.html
.....

but in the contents I can see they are correctly downloaded.

In the generated Index.html file I can see that the urls are pointing from the
original name to the wrong name.

A real example from an Index.html :
 9:41 AM       737941 {PDATS800.msa}
wherein the PDATS800.msa points to PDATS800.html
and in the logfile it says
10:55:12 Warning:  File not parsed, looks like binary:
site/structure/PDATS800.msa

And yes, the file is binary

To be clear : 
- it's not happening for all files, seems to be random and when not repeatable
for the same files
- I have similar problems on different sites

Am I the only one encountering this kind of problems ?
What's going on, what might be the reason, what can been done ?
Thanks