Lost LOD settings Majestic Falls V2 Going back to 1954

[render2017]

I am developing a route based on Mark Howarth's classic Majestic Falls.
I have Mark's full support and he has been helping me immensely.
The problem is that in the route I have suddenly seen structures starting to disappear. I checked the details and note that many have different LOD than originally specified. They are random numbers from 1 to 6. Edit them back to 10 and all is good.
Can anyone suggest
a. How can can stop this going further
b. How to fix it short of editing every structure/scenery element individually
If you would like details of the route which does include most of the majestic Falls route plus some 15 miles extra mainline and several branch extensions you can view my YouTube video on this on the video board .
Alternatively, search YouTube for "Train Simulator 2017 - Majestic Falls V2 Going back to the 1950’s"
I have included a full detail topic in the video posts.
Thanks hope someone knows what is going on!
Cheers
render2017

[Chacal]

You can use RW-Tools to do a global search-and-replace of the relevant part of the blueprints.

[render2017]

After testing, I note that LOD is not set in concrete, it seems TS2017 randomly changes LOD at startup. I added a row of items making sure they were all at least LOD 7, and when I closed and re-opened the route each item in the row had a different LOD. I checked this in TS2015, and it did not do this, seems a new bug which of course, now will never be fixed. It seems neglectful that Dovetail have finished with TS2017 development but still sell it (make money) with no obligation to make it work right for new purchasers.

[buzz456]

I must say that I have never had that problem.

[render2017]

One possible theory is that I do not run with viewing distance set at maximum. It could be this affects some things, but only in routes being developed/edited. Not proven I am grasping at straws.
Theory two . . . I often copy items to multiply instances, I wonder if the copy command has some small issue with LOD . . . also unproven.
Sigh ! ! !
Colin R