That's a can of worms you're opening here. I'd say the jury's still out.
As far as font size is concerned, this study (
pdf) concludes that
there were no significant differences (for sizes 6-16) in reading
performance or accuracy due to either passage length or age there was
variation in subjects’ preferences on the text sizes used.
They compared a group of age 18-29 against one of 61-78 and used PDAs as test material. Predictably, older people prefer larger type, but they seem to be as accurate in reading it as youngsters are.
On the contrary, this 2007 study (
pdf), involving vision-impaired readers, states a preference for sans serif typefaces between 16 and 18 points, but fails to draw any hard conclusions or advice for typography for the 'general', non-impaired populace.
A collection of academic evidence articles can be found here, on a blog that is dedicated to the research into legibility as a function of typeface. Looks like a lot of articles are worth the attention. The blogger himself seems to draw the conclusion that there is no 'ideal' or 'most legible' typeface, as per his article on the blog's front page.
Another extensive literature study on the sans vs. serif debate doesn't find any significant difference, and seems well worth the read as well.
Moreover, there's more factors to consider than just type size and whether the typeface has a serif or not. Variables like relative x-height, openness of the counters, leading, tracking and font weight are making this even more of a wasps' nest.
All this is just a half hour's worth of Googling and scanning, and it seems I haven't glimpsed the bottom of the rabbit hole yet. If anything, I'll draw the conclusion that your question deserves a close vote for 'too broad'...