Fixes: esp8266#8938

Also increase accuracy of mapping and fixing map issues with large values.
Still only using integer operations.

Test function to see absolute error compared to floating point map function using `double` precision floating point values.

```c++
# include <iostream>
# include <string>

long test_map(long x, long in_min, long in_max, long out_min, long out_max) {
    const double out_length = out_max - out_min;
    const double in_length  = in_max - in_min;

    if (in_length == 0) { return in_min; }

    if (out_length == 0) { return out_min; }

    const double delta = x - in_min;
    const double value = (delta * out_length + (in_length / 2)) / in_length + out_min;
//    return std::round(value);
    return value;
}

long check(long x, long in_min, long in_max, long out_min, long out_max) {
    const long floatvalue = test_map(x, in_min, in_max, out_min, out_max);
    const long map_value = map(x, in_min, in_max, out_min, out_max);
    return floatvalue - map_value;
}

int main()
{
    long input = 0;

    for (size_t i = 0; i < 15; ++i) {
        input = (i / 3) * 2500 - 1 + i % 3;
        const long values[] = {
            check(input, 0,        1000,   255,      0),
            check(input, 0,        1000,   0,        255),
            check(input, 10000,    0,      100,      10100),
            check(input, 10000,    0,      10100,    100),
            check(input, 10000,    0,      0,        1024),
            check(input, 10000,    0,      1024,     0),
            check(input, 1000,     0,      0,        10240),
            check(input, 1000,     0,      10240,    0),
            check(input, 0,        1000,   0,        10240),
            check(input, 0,        1000,   10240,    0),
            check(input, 0,        10000,  10240,    0),
            check(input, 10234567, -12345, 10234567, 100),
            check(input, 10000,    0,      10234567, 0),
            check(input, 50,       1,      10234567, 0)
        };
        std::cout << input << ":";
        constexpr size_t nrvalues = sizeof(values) / sizeof(values[0]);

        for (size_t i = 0; i < nrvalues; ++i) {
            std::cout << "\t" << values[i];
        }
        std::cout << "\n";
    }
}
```

Output:
```
-1:	0	-1	0	0	0	-1	0	2	2	0	0	-1	2	2
0:	0	0	0	0	0	0	0	0	0	0	0	-1	0	1
1:	-1	0	0	0	-1	0	1	0	0	1	1	-1	0	0
2499:	1	-1	0	0	0	0	1	1	1	1	0	0	0	0
2500:	1	1	0	0	0	0	1	0	0	1	0	0	1	0
2501:	0	0	0	0	0	0	2	0	0	2	1	0	0	1
4999:	1	-1	0	0	0	0	1	1	1	1	0	0	0	0
5000:	1	0	0	0	0	0	1	0	0	1	0	0	1	1
5001:	0	0	0	0	0	0	2	0	0	2	1	0	1	0
7499:	1	-1	0	0	0	0	1	1	1	1	0	0	1	1
7500:	1	1	0	0	0	0	1	0	0	1	0	0	0	0
7501:	0	0	0	0	0	0	2	0	0	2	1	0	1	0
9999:	1	-1	0	0	0	-1	1	1	1	1	0	0	1	0
10000:	1	0	0	0	0	0	1	0	0	1	0	0	0	0
10001:	0	0	0	0	-1	0	2	0	0	2	2	0	0	1
```