When it comes to evaluating burn-in on an audio device, unless an individual has an in-use device and an equivalent in-design idling device, it is only recollection that is the guide to assess perceived changes to the end sound being evaluated. Recollection of a device's usage hours 1 - 10, is forgotten when evaluating the difference during usage hours 150 - 160. Reported evaluations are for me, taken with a 'pinch of salt'. The difference in the timeline is too great to produce a reasonable appraisal.
Easily exchangeable components, which are a strong influence on the produced end sound, will create an improved opportunity to evaluate where the memory relied on is of a much shorter period. For certain components used in a circuit, a minute is all that is needed to complete an exchange. In relation to the below, I have been part of sessions where these methods are used. Where Valves are concerned, I have experienced many exchanges of equivalent types in one session. Through having my own collection of close-measured valves of the same types, I am familiar with how a Power On Valve compares to a Valve that has been idling and then exchanged into the circuit. OP Amp's are components I have only heard in A/B exchanges. I have also been able to experience circuits designed to have the components used to undergo speedy exchanges on a single circuit. Also, I am familiar with a set up where there are two of the same devices, where one has a role to be the experimental unit and the other is the unchanging control unit.
In all experiences had, there is always something occurring that has an increased attraction, when a component has been in use for a period if time. Especially noticeable when comparisons are between components with much less usage period or presented cold. What I can not state with conviction is that a short use of a cold component does become close to the attraction perceived from the increased use component. Only at the immediate use of the circuit, after an exchange, is there something perceived as an audible difference. My experiences leaves me to suggest this perceived difference quickly diminishes as the exchange part is in use for a short period only.
Valves are being affected by a break-in period / exposed to a Power On period. It is easy to A/B evaluate between a Valve Powered On for a period and an idle Valve being added to the circuit as an exchange component. This is because having valves at hand with a known close test measurement between them makes exchanges speedy to achieve. Like-for-like exchanges are easy between a Valve having been exposed to a Power On and the one idling. If sound is to be influenced, it will be perceivable without much of a recollection involved.
A similar experience can be had between OP Amps of a design that are not attached within the circuit using solder. OP Amp's that are a pin engaged push connection are suitable to trial. Commonly, an Op Amp shares a recommendation of 'burn in' to approx' (100 - 200 hours). An OP Amp device is able to be powered on permanently, replaying music. Every 20-30 Hours a equivalent OP Amp with almost zero usage can be the exchange component. This is a realistic way to evaluate the effect of extended component usage on a circuit. I have not experienced this method, only exchanges of an alternate component.
Either of the above will give an immediate A/B evaluation of a device using a particular component only, not the overall device, but a Valve or OP Amp is quite an influence on the end sound being produced. Reports generated from such an experience will be more reliable that ones made from recollections over a much longer timeline.
More complicated are Capacitors and Resistors, these will need a crocodile clip in use within the circuit to enable the clip to attach to a components lead out wire. Only achievable when an improved skill set is at hand and needed to achieve this method. There are good reasons not to venture down this route, but learning is learning, and it is a practice seen used.
A Capacitor or Resistor is able to have an influence on the end sound. Learning how they generate sound, from the earliest period of usage, through to an extended period of use, will do much to identify the voicing developing for the devices circuit. In a very limited capacity, I have been able to compare identical circuits with a difference in their usage hours. An experimental and control device has been used for a Phonostage Build. This same build also allowed for experiencing exchanges of components in the circuit for equivalent parts, and also components with changes to the spec for the parts.
Equivalent parts exchanged in my assessment do not do much that is perceivable or are notable for impact on the circuits produced end sound. Changing the Spec' for a part, now this is different, this is where there is a notable change, where there is real magic to be found, and great impressions are able to be created.