Photozone tests of the 31 ltd. On Photozone and 10 MP
Max. resolution at ƒ4 2345 lw/ph. (Line widths per pixel hight) or how many distinct lines can the lens and sensor reproduce.
At 16 MP. 2596
The increase is about 250 lwph based on 6 more MP.
At 10 MP you have 243.5 lw/ph per MP.
At 16 MO you have 162.5 lw/ph per MP
So one thing that is happening is the law of diminishing returns. 50% more pixels... 15% more resolution.
Your 18-135 at 16 MP produces 2683 lw/ph @ 24mm so it isn't the lens.
You would expect the Pentax current focus on more capable lenses digital sensors is in part due to the lack of resolution to be gained from using a film era lens on a digital sensor as the sensor density increases.
How is it on Nikon?
At 16 MP Nikkor AF-S 24mm f/1.4 G ED = 2883 lw/ph
At 24 MP Nikkor AF-S 24mm f/1.4 G ED = 3704 lw/ph (tested on the D7200)
50% more MP, 28% improvement... twice as much as the 31 ltd. Modern glass vs. film era glass.
Looking at sensors, and these numbers are not comparable to the above numbers, testing sites each have their own quirks, so wipe those numbers out of your mind but just for comparison.... the D7200 tested in Imaging resources.
Using a Sigma 70 2.8 the 7200 tests at 2950 lw/ph
Using a Sigma 70 2.8 the K-3ii tests at 2750 lw/ph
The K-S2 tests at 2500 lw/ph
If memory serves me well, because I can't find the source, the K-5 (16 Mp) tested at about 2100 lw/ph.
Without breaking it all down, you can see that although the Sigma 70 is an excellent lens 4MP gets you 400 lw/ph from 16-20 and 250 MP from 20-24. Again diminishing returns rears it's ugly head.
One more thing to consider, your largest print is 8x10.
At 16 MP with a Sigma 70 you are 8 inches deep for 2100 lwph... that's 262 distinct lines per inch. .0038
At 24 MP you'd have a resolution of 344 lines per inch or a minimum line thickness of .0029 inches.
So do you think you can see a difference of .0009 inches?
The issue here is your 8x10s don't stress a 16 MP camera. At that size and 262 distinct lines per inch, you are very close to the maximum DPI of your printer. So there's not much left to be gained. With the only test I've seen on this saying most people can't tell the difference between 72 DPI and 150 DPI, you can begin to understand the problem with increasing detail.
Something that is often overlooked on the forum is how little you get for your money as sensors get better using older glass. But even then with the Nikon 24, you still only get a 28% improvement for 50% more MP.
And that will only be visible when you have detail in your image that uses the available resolution. For example, if the finest detail to be resolved in your image is 100 distinct lines per inch, and think about that, that's pretty good detail. The 262 distinct lines per inch is pretty darn good. The simple fact is, based on my own observations at craft shows etc. MP stopped being a critical component of a camera in terms of selling even 30x20 canvases at about 12 MP. Everything after that in most images is over kill and goes to oversampling. Especially in 8x10 prints where no one is really sure that the printers involved can actually resolve that kind of detail. We know your monitor can't.
I'm using a 2010 27 inch iMac for 36 MP files, and it became manageable when I upgraded to an SSD drive. But mine is an i7 quad core. The processor makes a huge difference.
Most people pay for a lot of capability they will never use.