There is something called sampling bias - or sampling error is another name.
Remember that insects in general do one of these: (A) egg -> larvae -> metamorphosis -> adult, (B) egg -> nymph (several stages) -> adult
Some things to consider:
The free flying bugs we see, especially type A, are often very short lived as adults. Most species require very specific habitats. Also, many live less than one or two days as adults. Put them in a jar and they die the next day. Leave them outside and they die the next day anyway. So maybe the jar is always not a cause of mortality. Like
@BillTre mentioned.
When you capture bugs in a jar, you are getting a very very tiny sample of all of the bugs in your local area. A one acre field can contain as many as several hundred million insects. There are about 200 million inivididual insects on Earth for each human.
Your sample does not represent all of the bugs in your yard or house. So 'all bugs die in glass jars' may not be accurate in that sense. It may be that more slow moving species are easier to catch, and happen to be type A. And these guys are not larvae which usually outnumber adults for most species.
Long ago I ran some bug traps in Peru. Light traps at night trapped so many bugs we could only run them for a few minutes before the collection bag (about 3 liters) was overflowing. A colleague of mine used bug bombs to collect insects in the Amazon basin. This was pre-1970. He put down a circular thin mesh mat around the base of a tree, fired off a bug pyrethrin(?) bomb, then watched insects rain down for 15 minutes. The mat was about 10m radius, and usually wound up with about a 5cm (2 inches) deep bug layer. Millions of insects. I do not think he ever counted them or would would still be doing the counts. He did weigh them, I think.
The point is your sample of bugs in your jar does not really represent what you think it does.
Example from Panama:
Bug species in one acre (not individual bugs):
https://www.wired.com/2012/12/panama-bug-count/
San Lorenzo Forest - where the bug count above was taken:
http://ufdc.ufl.edu/AA00022174/00001
A field left fallow for a few years also has large insect populations.