Six years ago, it could, with ease. Remember, the big problem for OCR has been the available processing power and that hasn’t been a problem for a while. You can avoid it with clever algorithms, but you can brute force it with more power.
Heck, in 1996, Apple had a commercial PDA that could read cursive writing - specifically mine, which humans have problems reading. The government had a “brute force” solution long before that.
All I can say is wow. Several years ago, nearly 10, where I worked we tried scanning specifications and reports. It took so long to proofread and make corrections, we hired some temps to type them directly into Word. This was software allegedly designed to read typewritten text and it was not much help. I don’t remember the software name; the IT people bought it. For all I know it was The Billy Bob Sofware Company of Bulls Gap, Tennessee.