I strongly disagree with the statement that there is no reason for electronic system to behave differently after burn-in.
I spent 25 years as a SysAdmin [later as a Network Architect, last 15 years as a Software Engineer]; and when installing new hardware, standard procedure was [and still is] to burn in systems for a minimum of 72 hours - and usually 168 [yes, a full week] ... for servers, this meant running low level memory and disk array diagnostics continuously.
The time spent up front saved me quite a lot of time down the road because the vast majority of the time [better than 98%, yes I/we kept stats], if a piece of hardware was going to misbehave it did so during the burn-in process rather than after it had been put in to production.
This was true for servers, disks/arrays, routers and switches, and for UPS/power conditioning hardware that kept us insulated from problems on the grid.
Just my $0.02 + more than 40 years of practical experience, your mileage may vary ....
-dmm