I am well aware of these two failure deficiencies, as they are two of the only areas where I have no taken complete control over the systems to achieve my desired level of accuracy.
The former, I have not thought of a good way to do this without overriding the native autopilot, which becomes a near impossibility when dealing with so many different 3rd party developers for GPS units, who also override the native autopilot for their own purposes.
The latter is only slightly difficult, because MSFS does not provide a persistent indicated airspeed value (one not affected by the simulator's own sense of icing and blockages, which cause it to go to zero...), and I would worry about trusting my own derived value (that I currently use for sounds) for something as critical as airspeed.
I have wanted to tackle the entire genre of pitot/static failures for a while, but the math is fairly difficult to get them real world accurate with all lapse rate and temperature effects (high to low, look out below), and I know that is the first thing people would want to try if I make that kind of realism claim.
In short on both issues: It's never as easy as it seems, and sometimes it's an order of magnitude harder than you might even imagine.