And also, I'd like to point your attention to Aniseikonia, an EXTREMELY underdiagnosed issue:
http://www.opticaldiagnostics.com/info/aniseikonia.html
Neither clinical neuro-ophthalmology, nor orthoptists, nor optometricians, nor opticians, commonly have the equipment to test this, let alone know about it. (I had to literally have a lens kit dusted off in my case. One that nobody besides one entire person in the city knew about besides me.)
And the fact that one must consider the current distance and measuring equipment used for vision tests insufficient:
http://www.tedmontgomery.com/the_eye/acuity.html
To quote:
"Traditionally, optical infinity has been accepted to be 20 feet or, approximately, 6 meters. However, at this distance, there is an accommodative demand on the eye of about 1/6 D (one-sixth of a diopter). This amount of accommodative demand can be significant for some people. For very discriminating observers (such as myself), an accommodative fluctuation during an eye examination of more than 1/8 D can result in a variable endpoint in measuring a person’s refractive error (resulting in an imprecise lens prescription), and 1/6 D is even greater than 1/8 D.
As a result, it is recommended that the viewing distance (d) in an examination room should be great enough to create no more than a 1/8 D accommodative demand on any patient’s eyes. I maintain, then, that optical infinity, for purposes of examining the refractive error of the human eye, is at least 8 meters or 26¼ feet, rather than merely 6 meters or 20 feet."
This combines with the fact that that the human eye experiences "accommodative microfluctuations" which have a range of about 0,5 diopters (aka ±0,25 diopter), and it does so even at true optical infinity.
This then makes any assessment done without cycloplegia problematic, but assessment under cycloplegia also has its issues:
There exists a dissertation from a German Technical University (Well, "Fachhochschule", look the term up if you have an obsession with detail like I do) that beyond the shadow of a doubt proves that Schack-Hartmann sensor based Wavefront aberrometers can and will do better than humans when measuring the ideal prescription for a human eye. However, this dissertation, even as good as it was (they had a huge sample size and accounted for damn near everything), has several limitations and also points several problems out, and in addition to these, I'll add a few ones I've observed myself and that one can find in the literature (this is from memory and reading a lot of the literature, but keep in mind I can't find the original paper [in German] anymore, and some of this comes from other papers):
1. The default setting of most wavefront aberrometers defaults to 3 eye measurement cycles per eye measurement. However, the study found quite clearly that at this setting, a wavefront aberrometers will very often do worse than an experienced human optometrist or orthoptist, even if they do account for the distance problem I described above, which nearly all of them don't. A 5 measurement cycles however elevate the wavefront
2. Most wavefront aberrometers lack the capability to do a proper measurement of a fully dilated pupil as they expect a maximum pupil dilation of about 5mm, which is ridiculously tiny
3. Wavefront aberrometers cannot compensate for deeply (or at all) cramped cilliary muscles, an extremely common issue
4a. To get a dilated pupil, one must apply a mydriasic (i.e. pupil dilating) agent such as cyclopentolate.
4b. To get a relaxed cilliary muscle, one must apply a cycloplegic (i.e. cilliary mucscle paralyzing) agent such as cyclopentolate.
5. Most wavefront aberrometer measurements occur in the office of an optician, who legally may not dispense cycloplegic & mydriasic eye drops (such as cyclopentolate).
6. This is a good thing, because cyclopentolate is an muscarinic antagonist. Normally relatively harmless, when given to people with various atypical neurological features such as for example some of (but not all of) those exhibited by some of (but not all of those) those diagnosed with schizophrenia, it can lead to acute psychosis.
7. The above seems doubly unfortunate because schizophrenia seems to have an - extremely little researched - link to the previously mentioned Aniseikonia: http://www.schizo-binoc.de/ [Personally, I think some of the hypotheses on the neurological mechanisms of this link as offered by Hildegard Korn seem whack, but if you look at purely her quite empirical - albeit possibly anecdotal - data, you'll see that the pattern she describes along with the formal errors in commonly used diagnostical procedures she outlines quite flawlessly do add up to something and could probably get easily explained, in terms of neurological mechanisms behind the observed biophysics, by a more respectable theory on the neurological mechanisms behind it.)
And then there's the whole patent lockdown between all the big glasses manufacturers, but I cba to continue typing this out.
I truly despise the optical and ophthalmological industry. They still haven't heard Sue Barry's message although she's shouting it with all the force she can muster.