This might be a naive question but hear me out, can't we just connect the two wires in parallel to build an OR gate? Like this:
simulate this circuit – Schematic created using CircuitLab
Attach maybe a resistor at the end to have a consistent value if needed.
There's some excellent discussion here but, to summarize it a bit, aside from the "shorting inputs together" part, using an OR gate gives you output drive capability, and the possibility of some level of ESD protection, as well as better-defined transitions for indeterminate or slow inputs, all without the several-hundred mV diode drop.
The output of an OR-gate is HIGH, if just one of the inputs are HIGH. So the truth table is
$$\begin{smallmatrix}\begin{array}{rr|c} I_1 & I_2 & O\\ \hline 0 & 0 & 0\\ 0 & 1 & 1\\ 1 & 0 & 1\\ 1 & 1 & 1 \end{array}\end{smallmatrix}$$
In your diagram, you have shorted \$I_1\$ and \$I_2\$. This means that \$I_1\$ and \$I_2\$ will be at the same potential. But what if you force \$I_1\$ to be HIGH and \$I_2\$ to be LOW? You are forcing them to be at a different potential, but they have to be at the same potential... so we have a contradiction.
Even if we look past this, which input will dictate the output? If \$I_1\$ is HIGH and \$I_2\$ is LOW how can you possibly conclude what the output will be? You cannot combine multiple signals into a single node.
