fix: ensure that untyped values are converted to the same type

[secDre4mer]

This fixes an issue where two untyped values with different itype could not be operated on.

Minimal example to reproduce this:

package main

import "math"

func main() {
	println(math.MaxFloat64 - 1.0)
}

Result: invalid operation: mismatched types untyped float64 and untyped float, but should be 1.7976931348623157e+308

[CLAassistant]

All committers have signed the CLA.

[mvertes]

Looks good to me.
Thank you for this contribution.