Thanks for asking!
Depends on the parameter, and it depends on what one means by trust. Also, the apex is not a new it thing - its been around a while now and has been put thought the paces by people I trust, not just me. I had that same mindset you mention more when it was new.
I have explained publicaly many times, and this is in at least one of the MACNA talks which specifically talks about apex, I generally use testing to look for for trends not for particular numbers.
Temp is pretty straight forward, yet I still run other temp devices for comparison and ballpark confirmation (I once tested bit temp digital thermeters and out of the box some of them read 3 degrees f different from each other!). The trident I have had since last November I think and I have done enough comparative testing (hobby and more professional) to make me trust the trends, yet I still cross check it with other methods from time to time - I also worked with the prototype for over a year on alk and that was rock solid. Even the salinity probe, which goes goody from time to time generally gives a good trendline, I cross check with a misco digital Refracto (based on a lot of comparative testing with solutions of known value over time).
I have two issues with the MSM and have been testing it for like a year, so my feelings about it are different from someone that is just getting it. The trends seem to follow other testing method trends, but some of the numbers seem too far out of range of other methods for me (other have it better). The other is that there is no way to directly validate the readings from the msm with a known solution. I think they are closing in on making it work and look forward to the day I can trust it more - but as I say, the trends seem good.
For me, understanding the realities of the numbers and how they are produced is important. If a number comes up goofy, i look at that parameter more closely before taking action.
Please let me know if any of that doesn’t make sense - it’s a lazy Sunday for me. Thanks!